compaction, rock properties and avo modeling in the … · finds is progressively decreasing. as...

Master Thesis in Geosciences

COMPACTION, ROCK

PROPERTIES AND AVO

MODELING IN THE GOLIAT FIELD,

SW BARENTS SEA

A petrophysical approach

Honore Dzekamelive Yenwongfai

COMPACTION, ROCK PROPERTIES

AND AVO MODELING IN THE GOLIAT

FIELD, SW BARENTS SEA

A petrophysical approach

Honore Dzekamelive Yenwongfai

Master Thesis in Geosciences

Discipline: Petroleum Geology and Petroleum Geophysics

Department of Geosciences

Faculty of Mathematics and Natural Sciences

UNIVERSITY OF OSLO

[01.06.11]

© Honore Dzekamelive Yenwongfai, 2011

Tutor(s): Nazmul Mondol (UiO) and Jens Jahren (UiO)

This work is published digitally through DUO – Digitale Utgivelser ved UiO

http://www.duo.uio.no

It is also catalogued in BIBSYS (http://www.bibsys.no/english)

All rights reserved. No part of this publication may be reproduced or transmitted, in any form or by any means,

without permission.

Key words: Goliat Field, Compaction, AVO, Exhumation, Synthetic seismic

http://www.duo.uio.no/

http://www.bibsys.no/english

Preface

i

PREFACE

This thesis is part of the BarRock project and is submitted to the Section of Petroleum

Geology and Petroleum Geophysics (PEGG), Department of Geosciences, University of Oslo

(UiO) in candidacy of the, M.Sc. in Petroleum Geology and Geophysics.

This research has been performed at the Department of Geosciences, UiO during the period

from January – May 2011 under the supervision of Associate Professors, Nazmul Mondol and

Jens Jahren of the Department of Geosciences, University of Oslo, Norway.

Dedication

ii

DEDICATION

To Shufai wo Buea, Ma Ndi, Njingti’s, Ngala’s, Emesum’s, Emma, Fredo, Leo and the entire

Fai’s family.

Acknowledgements

iii

ACKNOWLEDGEMENTS

It would have been next to impossible to write this thesis without my supervisors, Associate

professors Nazmul Mondol and Jens Jahren whose help, guidance, encouragement, and

supervision from preliminary to concluding levels, enabled me acquire and develop new

analytical techniques.

I am grateful to all my lecturers in the section of Petroleum geology and Petroleum

geophysics, most especially, Johan Peter, Knut Bjørlykke, Faleide Inge, Roy Gabrielson, Leiv

Gelius, Dag Karlsen, and Michael Heeremans for their support, constructive feedback and

recommendations. I would always be indebted to you all.

Many thanks goes to my course mates in the Petroleum Geology and Geophysics who make

up a significant diversity from different continents, and enabled me develop alternative

models of thinking and an open minded culture.

I also want to acknowledge my study group mates, Abel Onana, Agus Fitriyanto, and

Piratheeben K, who have stood by me all the way. Your team spirit, academic and social input

will forever be missed. I want to convey thanks to Fawad Manzar for his constructive

feedback and relevant discussions in this work.

I am grateful to the Emesum family, whose friendship, hospitality and love, enlightened and

helped me stay focused. The good memories and times well spent will always be

remembered.

Finally I want to thank all my friends, espacially Olivier Pamen, Daniel Ngembus, Ashu

Mpame, Sone Brice, Geraldine Njumbe, Aghendia Alemngu, and last but not least Pongwe

Fadimatou. Your moral, emotional, and spiritual support throughout my studies will forever

be cherished.

Abstract

iv

ABSTRACT

The PDO approved Goliat Field, situated on the Finnmark Platform in SW Barents Sea region

represents the first oil field to be developed on the Norwegian Barents shelf. Cenozoic

exhumation still poses significant challenges in developing frontier exploration models due to

its effects on the different elements in a petroleum system.

Understanding the nature of the transition from mechanical to chemical compaction, the

degree of overconsolidation and AVO modeling of selected reservoir intervals is the main

theme of this work. Data from six wells, laboratory compaction curves and published shale

compaction trends have been used to evaluate the rock properties as a function of depth. AVO

synthetic single interface models have been carried out using Hampson-Russel.

Experimental laboratory compaction data coupled with compaction trends in the Goliat Field

indicate that the amount of exhumation ranges between 700 – 1500m. An integration of the

Vp-depth trend together with a porosity-shear modulus cross plot, show that the transition

from mechanical to chemical compaction for siliciclastic rocks occurs at approximately 600m

BSF. This temperature controlled transition represents a silica phase transformation. The

Chemical compaction trends show a smaller change in Vp with depth compared to the

mechanical compaction domain. Velocity inversion with depth due to the presence of source

rocks, effect of pore fluid and pore pressure has been demonstrated for different wells.

AVO modeling for different fluid saturation scenarios indicates that the synthetic seismic is

sensitive only to the initial 10% gas saturation in oil – gas system, with the saturated bulk

modulus being the main controlling parameter. The insitu AVO response for the Tubåen

reservoir is class IV, meanwhile the Kobbe reservoir gives a class III signature. The

corresponding gas models indicate that, ∆Vs and Poisson’s ratio are key parameters

increasing the reflectivity with offset (angle), meanwhile the impedance contrast in the half

space single layer models determines the magnitude of the reflection coefficient in both the

Tubåen and Kobbe reservoirs.

This study demonstrates that AVO modeling can be used for fluid prediction ahead of drilling

during exploration and reservoir monitoring during production. Uplift estimates can be used

to correct the porosity depth relationships used in reservoir characterization work flows, and

also in assessing the degree of tertiary migration from traps due to exsolution of gas.

Table of contents

v

TABLE OF CONTENTS

Preface ................................................................................................................................i

Dedication.......................................................................................................................... ii

Acknowledgements.......................................................................................................... iii

Abstract..............................................................................................................................iv

Table of Contents................................................................................................................v

List of Figures...................................................................................................................viii

List of Tables.....................................................................................................................xii

CHAPTER 1 GENERAL INTRODUCTION

1.1 General.................................................................................................................................1

1.2 Exploration history in the Barents Sea area ........................................................................2

1.3 Goliat field ..........................................................................................................................3

1.4 Research Objectives ............................................................................................................5

1.5 Database and Methodology ................................................................................................6

1.6 Chapter description .............................................................................................................8

CHAPTER 2 REGIONAL GEOLOGIC SETTING

2. I Structure and Tectonic ....................................................................................................... 9

2.2 Stratigraphy ........................................................................................................................11

2.2.1 Kapp Toscana Group.......................................................................................................14

2.2.2 Sassendalen Group..........................................................................................................15

2.3 Petroleum system ..........................................................................................................….15

2.3.1 Source Rock ................................................................................................................... 17

Table of contents

vi

2.3.2 Reservoir units ................................................................................................................18

2.3.1 Trap .................................................................................................................................21

2.4 Exploration challenges on the Barents shelf ......................................................................24

CHAPTER 3 COMPACTION AND ROCK PROPERTIES

3.1 Introduction .......................................................................................................................25

3.2 Theoretical Background.....................................................................................................26

3.2.1 Mechanical Compaction .................................................................................................26

3.2.2 Chemical Compaction .....................................................................................................28

3.3 Materials and Methods....................................................................................................... 31

3.4 Results................................................................................................................................ 33

3.4.1 General porosity/density/Vp versus depth trends ...........................................................33

3.4.2 Vp-depth trend for well 7122/7-3 ...................................................................................36

3.4.3 Transition from mechanical to chemical compaction .....................................................37

3.3.4 Uplift estimation ............................................................................................................ 39

3.4.5 Sand and shale compaction trends ..................................................................................42

3.3.6 Effect of pore fluid and pore pressure ............................................................................44

3.3.7 Effect of source rock on Vp-depth trend..…..................................................................45

3.4 Discussion ..........................................................................................................................47

3.4.1 Relationship between porosity/ density/ Vp versus depth trends ...................................47

3.4.2: Uplift Estimation ...........................................................................................................50

3.4.3 Transition from Mechanical to chemical compaction ....................................................52

3.4.4 Variations in the Sand and shale compaction trends ......................................................55

3.4.4 Effect of pore fluid and pore pressure ............................................................................56

Table of contents

vii

3.3.5 Effect of source rock .......................................................................................................57

CHAPTER 4 AVO MODELING

4.1 Introduction ......................................................................................................................58

4.2 Theoretical Background.....................................................................................................59

4.2.1 Vp-Vs Relationships ......................................................................................................59

4.2.2 Gassmann fluid substitution ..........................................................................................60

4.2.3 Synthetic Seismogram ...................................................................................................61

4.2.3 Angle dependent reflection coefficient .........................................................................63

4.2.4 Classification of reservoir sands based on AVO ..........................................................64

4.3 Database and methodology ............................................................................................. 67

4.4 Results ..............................................................................................................................72

4.5 Sensitivity analysis ............................................................................................................72

4.5.1 Variations in half space models.......................................................................................75

4.5.2 Effect of block size variation on the AVO signature ......................................................79

4.5.3 Kobbe and Tubåen angle dependent reflectivity comparison ........................................ 80

4.6 Discussion ......................................................................................................................... 81

4.6.1 Sensitivity study ............................................................................................................. 81

4.6.2 Variation in half space models ........................................................................................84

4.7 Uncertainties in the modeled scenarios...............................................................................88

CHAPTER 5 SUMMARY AND CONCLUSION

5.1 Summary and conclusion................................................................................................... 90

REFERENCES........................................................................................................................ 93

APPENDIX .............................................................................................................................98

Chapter 1. General Introduction

1

CHAPTER 1 GENERAL INTRODUCTION

1.1 General

In recent years, as the easy to find conventional hydrocarbon reserves in the earth’s crust are

being exploited, the oil industries tend to search in more difficult terrains like the arctic and

much deeper waters to match the growing demand for fossil fuels. Also, the global demand

for fossil fuel continues to grow unfortunately at a time when the size of new hydrocarbon

finds is progressively decreasing. As such, exploration for oil and gas over time has advanced

from being qualitative to quantitative. Quantitative studies of the subsurface in general and

hydrocarbon fields in particular, require a lot of integrated data and analysis from geologists,

geophysicists, petrophysicists, and reservoir engineers.

Fueled by progressive technological advances and breakthroughs in the oil and gas industry,

the possible computing power has also followed suite such that reservoir characterization has

extended from deterministic to probabilistic. Accurate characterization requires a combination

of 3D and 4D seismic volume interpretations, seismic inversion and amplitude analyses, rock

physics and AVO (amplitude versus offset) analysis. In some cases neural networks are also

applied to create 3D volumes of petrophysical properties to model inter-well data, thus

establish and visualize the spatial variations in reservoir parameters (Goffrey, 2007). Earlier,

geophysical data were mainly used in exploration, and to a smaller extent in the development

of discoveries. In more recent times geophysical and petrophysical data is integrated in

reservoir characterization schemes, and serves as a link between geologic reservoir properties

(such as porosity, sorting, clay content, lithology and saturation) and seismic properties (like

P-wave and S-wave velocities (Vp/Vs ratio), acoustic impedance, elastic moduli, bulk

density) (Avseth et al., 2010). Reservoir characterization therefore simply refers to

quantitatively assigning reservoir properties which usually show a non-uniform and non-

linear spatial distribution. As a consequence of this reservoir heterogeneity, effective rather

than absolute quantitative reservoir parameters become much more important as input for

reservoir simulations and subsequent optimal field development (Mohaghegh et al., 1996). By

applying an integrated approach in reservoir characterization together with geostatistics

(because a simple average of data from even closely spaced wells may lead to a misleading

analysis), an adequately constrained reservoir model will be the end result, which can then be


2

used to quantify the hydrocarbons in place and also in optimizing hydrocarbon production

(Jarvis, 2006).

This study will focus on an integrated approach to reservoir characterization of the Goliat

field, located on the sub area Finnmark West close to the Hammerfest Basin in the

southwestern part of the Barents Sea (Norwegian sector block 7122). A suite of quality

controlled well logs will be used to infer the rock property variations with depth in the study

area. Well logs will then be used to build high frequency synthetic seismograms to perform

AVO modeling.

1.2 Exploration history in the Barents Sea area

In terms of surface area, the Barents sea (245,000 Km2), is significantly larger than the

Norwegian North Sea (130,000 Km2) (Figure 1.1). Despite its size, the success rate in the

Norwegian Barents Sea (one in three wells) is higher than the Norwegian North Sea, with

respect to the number of dry wells drilled before the first commercial discovery (Ohm and

Karlsen, 2008). A number of 32 dry wells preceded the Ekofisk hydrocarbon field in the

Norwegian North Sea. Meanwhile more than 100 dry wells were encountered prior to the

Dutch Groningen discovery in the southern part of the North Sea. Approximately one out of

three wells generally has been successful in the Barents Sea (Ohm and Karlsen, 2008). Also

multiple source rock intervals at different stratigraphic intervals and have been documented in

the Barents Sea unlike the North Sea with just one major source rock (Kimmeridge shale)

which is equivalent to the Hekkingen Formation. Previously the most common play models in

the Barents Sea region involved mainly gas prospects such as the Snøhvit. The hydrocarbon

products are usually ranked to be less commercial because of predominance of gas over oil

and the distance to the gas market is also an issue. The predominance of gas over oil has been

attributed to uplift in this area. However oil discoveries such as the Goliat and Nucula, has

implications for established exploration models in this area. These models need to take into

consideration long distance migration as a result of different episodes of uplift and also the

dynamic nature of traps. A more recent significant oil discovery in Skrugard prospect (1250m

BSF in well 7220/8-1), has been made by Statoil ASA, and partners Eni Norway and Petoro

in April 2011. This discovery is located about 100 Km North of the Snøhvit gas field and has

an estimated 150-250 million recoverable barrels of oil equivalent (Eni Norway Goliat


3

Factpage). This discovery represents an important break-through for frontier exploration

activities in this area.

Fig. 1.1 Barents Sea exploration activity.

1.3 Goliat field

The Goliat field is a PDO (Plan for Development and Operation) approved field located in

block 7122/7 and 7122/8 (Figure 1.2) in the Norwegian sector of the Barents Sea (Production

Licence 229 awarded in 1997). It is about 50 km southeast of the Snøhvit field in the sub area

Finnmark West and about 85 km northwest of Hammerfest. It is the first oil field to be

developed in the Norwegian sector of the Barents Sea. The licensees of PL229/229B are

partitioned between two equity partners; Eni Norge (operator) has 65% meanwhile Statoil

Petroleum AS has 35% of equity interest in Goliat field (NPD Factpages).


4

Fig. 1.2 Location map of Goliat Field (NPD Factpages).

Unlike most discoveries in the Barents Sea which are rather gas prone than oil prone, the

Goliat field is mainly a crude oil field with a relatively smaller volume of natural gas. This

discovery came to light from the first exploration well (7122/7-1) in 2000. The water depths

in this area between 360 – 420m (Eni Norway Goliat Factpage). The reservoir lies at about

1100m. A total of 6 wells have been drilled in this area with one well 7122/7-5 being a dry

well among the other success stories. The two main reservoir intervals in this field are the

Kobbe Formation (Sassendalen Group) of Triassic age and the Tubåen and Fruholmen

Formations in the Realgrunnen Subgroup (Kapp Toscana Group) of Jurassic age. Both are oil

discoveries with an additional thin gas cap. A couple of other discoveries have been

documented in the Snadd Formation (Upper Triassic) and the Klappmyss Formation (Lower

Triassic). The main reservoirs in this field are located within the Triassic (NPD Factpages).

The development phase of this field started in the last quarter of 2009, using the Sevan FPSO

(Floating Production, Storage and Offloading) 1000 concept. This field is to be developed


5

with eight integrated subsea templates and about 32 well slots tied to the circular FPSO

system (NPD Factpages). Production is estimated to begin by the last quarter of 2013. The

pressure in most of the reservoir intervals is low, for example 123 bars have been documented

for Realgrunnen Subgroup and 192 bars for Kobbe Formation (Eni Norway Goliat Factpage).

As a result, this field will be produced with water injection to optimize recovery.

During the early phases of production, associated gas will be re-injected into the Kobbe

Formation until export through the Snøhvit pipeline to Melkøya is possible (NPD Factpages).

The low pressures in the reservoir represent a positive element for well control but on the

other side of the coin, more artificial energy input in terms of water flooding is necessary to

maintain reservoir pressures during production and thus optimize recovery. There is a

relatively small support from a natural gas drive due to the thin gas cap. The anticipated

production profile indicates a build-up to 5.4 million Sm3/year by the second year of

production followed by a relatively rapid decline to 1.7 million Sm3/year, then a steady

reduction to 0.5 million Sm3/year (Eni Norway Goliat Factpage). It is anticipated that the

maximum volume of gas production including re-injection will take place one year after the

onset of production. This volume is estimated at 1300 million Sm3

/ year. Goliat is expected to

produce for a period of about 15-20 years. Technological advances coupled with any

additional discoveries within its vicinity will probably be important in extending the life of

this field.

1.4 Research Objectives

The aim of this thesis is to set an initial framework for compaction and AVO modeling in the

Goliat oil field. This involves qualitative and quantitative description of the lithologic and

fluid properties of several effective reservoir intervals in the uplifted Goliat field. Emphasis is

laid on the shallower, thin Early Jurassic Tubåen reservoir and the deeper and thicker Mid-

Triassic Kobbe reservoir. An integrated study is therefore paramount to these outlined

objectives;

1. Assess the effect of lithology, pore fluid and pore pressure on the general rock

compaction trend.


6

2. Quantify the total amount of Cenozoic uplift affecting the Goliat oil field, and

assess the depth and nature of the transition from mechanical to chemical

compaction for the siliciclastic rocks.

3. Fluid prediction and substitution in determined reservoir intervals by

theoretically replacing the insitu fluid phase and monitoring the corresponding

changes in the synthetic seismic response.

4. Rock physics AVO modeling and classification of specific reservoir intervals.

1.5 Database and Methodology

This research employs an integrated approach, combining a petrophysical study from well

data and synthetic seismic data, to qualitatively and quantitatively determine reservoir

properties of the Goliat field.

Table 1.1 Well data and status (modified from NPD Factpages).

Well Entry date Completion date Purpose Content

7122/7-1 16.09.2000 05.10.2000 W Oil

7122/7-2 12.09.2001 19.10.2001 W / A Oil

7122/7-3 24.10.2005 08.01.2006 W / A Oil/Gas

7122/7-4S 21.09.2006 25.11.2006 W / A Oil/Gas

7122/7-5A 23.12.2006 13.01.2007 W / A Oil

The data used throughout this work for analysis is based on a complete suite of 6 wells,

drilled through Goliat field (Figure 1.3). Most of these wells are wildcat (w) and appraisal

wells (A) and one well being a dry well (7122/7-5). The wildcat and appraisal wells used

include; 7122/7-1, 7122/7-2, 7122/7-3, 7122/7-4S, and 7122/7-5 A. With production

scheduled to commence in 2013, the status of these 5 wells (Table 1.1) is now in the appraisal

and production phase.


7

Fig. 1.3 Well locations superimposed on the outline of the Goliat Field (NPD Factpages).

Data analysis and interpretation has been based on the Geoview, Elog, AVO modules in the

Hampson-Russel software packages. Elog module has been used for well log conditioning and

AVO module was used to generate and extract amplitudes from synthetic seismic data.

AVO modeling for ‘in-situ’ and ‘what if’ scenarios has been carried out in this work. This has

been used to determine the expected AVO anomaly, based on some principles such as Biot-

Gassmann fluid substitution model.

In addition to the well database, laboratory mechanical compaction data for pure sandstones

(Etive sand), synthetic silt-clay mixtures and published compaction trends have been used for

the compaction study. A detail description of materials and methods is given in the relevant

chapter.


8

1.6 Chapter description

Chapter 2 will describe the regional structural and stratigraphic setting of the Barents shelf

area and also give an overview of the petroleum system in the Goliat oil field based on

published data.

Chapter 3 will focus on compaction, rock properties, and exhumation estimates in the study

area. A brief theoretical background, comprehensive methodology and available data set is

presented in this chapter. Major findings with regards to exhumation and compaction trends

are also discussed, along with the effects of pore fluids, pore pressure and source rocks.

AVO fluid replacement modeling and classification of some defined target reservoir intervals

will be the main focus in chapter 4. A theoretical framework and the assumptions used in

modeling the target reservoir zones are also outlined. A sensitivity analysis, variations in

AVO half space models, and effect of block size are also discussed in this chapter.

Chapter 5 will then provide a summary of the entire work and the major conclusions arrived

at from the results obtained from the available data set.

Chapter 2. Regional geologic setting

9

CHAPTER 2 REGIONAL GEOLOGIC SETTING

2. I Structure and Tectonic

The physiographic association of the Goliat Field in the Barents Sea together with other

known chains of circumpolar basins such as the Sverdrup Basin and Mackenzie Delta of

Canada, Western Siberia Basin, Mid-Norwegian Shelf and the North Sea amongst others,

makes this area of the globe particularly interesting for hydrocarbon exploration. The Barents

shelf is located on the north-western edge of the Eurasian plate, with average water depths of

about 300m, defining an area of about 1.3 million km2. According to Larsen et al. (1993), the

area open for hydrocarbon exploration in this area covers an area of about 230000km2 which

compared to the Norwegian sector of the North Sea is more than one and a half times its area.

Fig. 2.1 Map showing the Goliat Field and the Troms-Finnmark Fault Complex (modified

from NPD website). Bathymetric map modified from Jacobsson et al. (2008).

The Goliat field cuts across the Troms-Finnmark fault complex and sits on the sub area

Finnmark West in the southwestern part of the Barents Sea (Figure 2.1). The geologic history

of most sedimentary basins usually involves an interesting interplay between tectonic

subsidence, possible reactivation of older faults and sediment supply ultimately affecting the

basin infill and the South Western Barents shelf is no exception. Two major continental

collisions and resultant orogenies typify the early history of the Barents shelf. The Caledonian


10

Orogeny closing the Iapetus Ocean approximately 400 Ma and the Uralian Orogeny

representing one of the last collision elements in Permian-Triassic times creating the

supercontinent, Pangea. These major Orogenies dominate the basement substructure of this

area, and it is thought to have influenced later structural development (Glørstad-Clark et al.,

2010).

Subsequent to Orogenic episodes, the Barents Sea area became dominated by later periods of

extension, beginning with the collapse of the previous orogenic belts and progressive breakup

of the already established Pangea supercontinent during the Late Paleozoic and Mesozoic.

This resulted to a complex mosaic of platforms, structural highs and rift basins across the

Barents shelf (Johansen et al., 1993) as shown in Figure 2.2.

Fig. 2.2 Main Structural Elements in the Barents Sea (Faleide et al., 2008, Gabrielsen et al.,

1990, Gudlaugsson et al., 1998).

The colors shown in Figure 2.2, show the focus of tectonic activity through time in the

Western Barents Sea. The focus of tectonic activity is seen to progress towards the west over

geologic time. The Hammerfest Basin is shown to be the main area of tectonic activity in Late

Jurassic – Early Cretaceous times.


11

2.2 Stratigraphy

The Barents shelf stratigraphic succession shows two major distinct sets of lithologies at

different Periods in its geologic history. The Devonian, Carboniferous and Permian on the

Barents Shelf is dominated by carbonates with some evaporites. Meanwhile the Triassic and

younger sediments are mainly represented by clastic sediments (Figure 2.3). This correlates

well with the drifting of the entire Barents shelf from warmer paleolatitude of 20oN during the

Carboniferous to a colder paleolatitude of 55o N during the Triassic and finally to present day

75oN. This clearly demonstrates a combined tectonic and climatic influence on lithology, as

progressively more temperate conditions prevailed (Worsley et al., 1986). The Triassic in the

Western Barents sea area rather represents a relatively quiet period with svalbard inclusive ,in

contrast to the Northern and Southern Barents Sea Basins which were progressively forming

depocenters through significant subsidence in these regions (Riis et al., 2008). Acording in to

Mørk et al. (1989), relatively thick Triassic clastic units (Figure 2.4) are present throughout

the Barents Sea usually showing coarsening upward sequences thus indicative of

transgressive-regressive depostional cycles. Multiple source rock intervals have been

documented in the Barents sea stratigraphy, from Carboniferous to Cretaceous (Ohm and

Karlsen, 2008).

The Lower-Middle Jurassic interval in the study area (Figure 2.4) is represented mainly by

sandstones which extend throughout the Hammerfest Basin and possibly also covered the

Loppa High and Finnmark Platform, but probably exhumed and eroded during later tectonic

events (Glørstad-Clark et al., 2010).


12

Fig. 2.3 Regional stratigraphy of the Barents Shelf. The cored interval in the Nordkapp basin

are shown (modified after Bugge et al., 2002).

The stratigraphy of the Goliat Field on the Finnmark platform shows an incomplete

stratigraphic section. This is the result of the Pliocene-Pleistocene glaciations, which eroded

most of the Paleogene and Neogene stratigraphic units. This erosion is more significant from

the eastern Barents Sea, with shallow units like the Torsk and Kviting Formations present in

the Hammerfest basin, but eroded out in the Bjarmeland Platform, Nordkapp Basin in the east.

The deepest well in this field is well 7122/7-3 and goes as far deep as the Permian with the

oldest unit being the Tempelfjorden Group (Table 2.1). The complete stratigraphy

encountered in well 7122/7-3 is shown in Table 2.2. The focus in this study will mainly be on

the Mesozoic Sub-Era and of particular interest the Triassic and Early Jurassic Periods.


13

Fig. 2.4 Lithostratigraphy of the Triassic in the Western Barents Sea (modified after

Glørstad-Clark et al., 2010).

Table 2.1 Well bores and corresponding oldest Group and Formations penetrated.

Well bore Total Depth

[m] KB

Oldest Unit Age Kelly bushing

elevation KB

[m]

7122/7-1 1524 Snadd Fm Triassic 24

7122/7-2 1418 Snadd Fm Triassic 18

7122/7-3 2726 Tempelfjorden GP Permian 25

7122/7-4S 2550 Havert Fm Early Triassic 23

7122/7-5A 2186 Kobbe Fm Middle Triassic 23

The lithostratigraphic description of the Kapp Toscana Group and the Sassendalen group that

follow are based mainly on published data in the Hammerfest Basin and other closely related

structural basins (NPD Factpages).


14

Table 2.2 Formations and Groups encountered in well 7122/7-3 (NPD fact pages).

Top Depth m

(BSF)

Top Name

0 NORDLAND GP

266 NYGRUNNEN GP

266 KVITING FM

282 ADVENTDALEN GP

282 KOLMULE FM

497 KOLJE FM

592 KNURR FM

650 HEKKINGEN FM

705 FUGLEN FM

719 KAPP TOSCANA GP

719 TUBÅEN FM

812 SNADD FM

1440 SASSENDALEN GP

1440 KOBBE FM

1676 KLAPPMYSS FM

1844 HAVERT FM

2227 TEMPELFJORDEN GP

2.2.1 Kapp Toscana Group

This group is dominated by sandstones, siltstones and shales, with age ranging from Late

Triassic to Middle Jurassic (Ladinian to Bathonian) and it is exposed along the Tertiary fold-

thrust belt on Spitsbergen, Barentsøya, Edgeøya, Hopen, Kong Karls Land and Bjørnoya. It

also extends southwards across the Barents Sea Shelf to the Bjarmeland Platform, the

Hammerfest and Nordkapp basins. This group has been interpreted in general to have been

deposited in a nearshore deltaic environment characterized by shallow marine and coastal

reworking of fluviodeltaic and deltaic sediments (Mørk et al., 1982).


15

The main reservoir units of interest in the Kapp Toscana Group from the Goliat Field are

represented by the Tubåen, Fruholmen (Realgrunnen Sub Group), Snadd (Storfjorden Sub

Group) and Kobbe Formations.

2.2.2 Sassendalen Group

This group is dominated by shales and siltstones with subordinate sandstones and minor

carbonate intervals. This group is of Early and Middle Triassic and is exposed along the

Svalbard Tertiary fold belt, Barentsøya, Edgeøya, southwestern Nordaustlandet and Bjørnoya

continuing in the subsurface southwards in the Barents Sea shelf to the Hammerfest basin.

This group has been interpreted to represent coastal, deltaic to shallow shelf deposits in

Western Spitsbergen. This group is represented by a series of stacked transgressive-regressive

successions, each formation being initiated by a regionally significant transgression (Mørk et

al., 1989).

2.3 Petroleum system

A petroleum system takes into consideration a pod of active source rock and its genetically

related oil and gas accumulations. This includes all the geologic processes and elements

which are essential ingredients for a hydrocarbon accumulation to occur. The essential

elements include; source rock, reservoir rock, seal and overburden. The two main processes

that need to be present include; trap formation and generation-migration-accumulation of

hydrocarbons. All events and processes need to be placed correctly in time and space to

obtain a higher probability for the occurrence of a functioning petroleum system. It is this

interdependence of the different elements and processes to form a hydrocarbon accumulation

that makes it a system. The proven hydrocarbon reserves in the Goliat field are evidence of a

working petroleum system. This petroleum system, like others on the Barents shelf, has

suffered the effects of different episodes of Cenozoic exhumation and erosion. This triggered

tertiary migration from the available traps. Uplift and erosion also possibly raised the source

rock ‘kitchen’ to shallower depths thus reducing its potential for producing more

hydrocarbons. This has serious consequences for exploration within such a context. An

overview of different source and reservoir intervals in the Barents Sea region has been

adequately summarized in Figure 2.5 by Dore (1995).


16

Fig. 2.5 Major source and reservoir rocks in the Barents Sea area (adapted from Dore,

1995).


17

2.3.1 Source Rock

This refers to a rock rich in organic matter (kerogen), which will generate hydrocarbons if

exposed to sufficient burial temperatures in the sedimentary basin. The most widely

distributed source rock candidate in this area is the Hekkingen Formation which is composed

of dark organic rich shales. The Hekkingen Formation (Figure 2.6) in the Barents Sea is the

equivalent of the Kimmeridge marine clay formation in the North Sea. This unit was

deposited in anoxic deep marine conditions, as consequence of the local barriers to circulation

created by the Kimmerian movements (Dalland et al., 1988). The Hekkingen formation

belongs to the Adventdalen Group, deposited during the regional Bathonian/Callovian marine

transgression.

Fig. 2.6 Core description of the Hekkingen Formation (adapted from Bugge et al., 2002).

The Hekkingen Formation is the most prolific in terms of its TOC (Total Organic Carbon) and

hydrocarbon generative potential. However the Fuglen formation in the same study shows a

more marine dominated depositional environment than the Hekkingen formation. Other

potential source rock intervals discussed in are found in the Nordmela, Tubåen, Snadd, Kobbe

formations including some source potential in the Permian (Ohm and Karlsen, 2008).

According to Ohm and Karlsen (2008), most of the Triassic source rocks enter the oil

window, when the Hekkingen Formation is just early mature. Meanwhile the Triassic source

rocks enter the gas window when the Hekkingen Formation becomes oil mature. In addition,


18

it was tentatively concluded based on isotope geochemistry, that Permian and Carboniferous

source rocks, yield oils with the lightest isotope values, meanwhile the Triassic sourced

intervals, yield rather intermediate, Isotope values.

Salt has a high thermal conductivity and tends to locally pull up the isotherms. The presence

of salt in the Nordkapp Basin has enhanced the maturity of the Hekkingen Formation in that

area. The Hekkingen Formation on the other hand gives rise to oils with heavy isotope values.

However most of the oils found in the traps in the Goliat field have mixed isotope

geochemical signatures, due to the variety of source rocks present in the different stratigraphic

intervals. Uplift and erosion, has negative consequences for hydrocarbon generation from the

source rocks on the Barents shelf. However, evidence of non-cogenetic gas has been

documented in this area, indicating the presence of a live petroleum system in the area (Ohm

and Karlsen, 2008).

2.3.2 Reservoir units

The main reservoir units encountered in the well data in this study are the Tubåen, Fruholmen,

Snadd, Kobbe and Klappmyss Formations. During the deposition of the Adventdalen Group,

sandstone distribution was restricted to local highs and platforms. The Triassic interval is

important in the Goliat Field due to the reservoir intervals deposited during that period. The

corresponding Lower Upper-Triassic (Carnian) reservoir units further north had a volcanic

provenance area. This makes them relatively texturally and mineralogically immature and

hence more susceptible to diagenetic alterations at depth, compared to areas further south.

Areas further south such as the Finnmark platform, were under the influence of continues

progradation from the Baltic Shield which represents a relatively mature provenance, with

primary reservoir quality increased by marine reworking in near-coastal environments during

highstand (Worsley, 2008).

The lithostratigraphic descriptions of these reservoir units are based on Dalland et al. (1988).

Tubåen Formation; this formation belongs to the Realgrunnen Sub Group of the Kap

Toscana Group and is dominated by sandstones with subordinate shales and minor

coals. It shows a tripartite development, with a shale interval sandwiched between

sand-rich units. This unit is probably distributed sub parallel to the Troms-Finnmark

Fault Complex. The base of this formation is of Late Rhaetian to early Hettangian


19

though it probably extends locally into the Sinemurian. The sand intervals in the

Tubåen Formation have been suggested to represent a stacked series of high energy

marginal marine (tidal inlet dominated barrier complex and/or estuarine)

environments.

Fruholmen Formation; this formation belongs to the Realgrunnen Sub Group of the

Kap Toscana Group. The base of this unit is early Norian while the top is thought to

be rather diachronous at the Triassic-Jurassic transition. Open marine shale intervals in

this formation grade to coastal and fluvial dominated sandstone sequences (Figure

2.7). This formation represents a depocentre in south with northward fluviodeltaic

progradation.

Fig. 2.7 Core description of the Fruholmen Formation (adapted from Bugge et al., 2002).

Snadd Formation; this formation belongs to the Storfjorden Sub Group of the Kap

Toscana Group and it is of a Ladinian to early Norian age. This unit generally shows a

coarsening upward succession from basal grey shales, into shales with interbeds of

grey siltstones and sandstones (Figure 2.8). The lower and middle portions of this

formation commonly have limestone and calcareous interbeds. This unit represents a

distal marine environment, following a regional marine transgressive event, during

which all structural highs and platform areas in the region became submerged.


20

Fig. 2.8 Core description of the Snadd Formation (adapted from Bugge et al., 2002).

Kobbe Formation; this formation belongs to the Sassendalen Group and consists of a

thick basal shale unit grading upward into interbedded shale, siltstone and carbonate

cemented sandstone (Figure 2.9). A coarser proximal facies development is typical of

this unit along the Southern margin of the Hammerfest Basin. This formation thickens

on the Troms-Finnmark platform, and is of Anisian age. A transgressive pulse marks

the base of this formation following a renewed clastic marginal marine regime from

the southern coastal areas.

Fig. 2.9 Core description of the Kobbe Formation (adapted from Bugge et al., 2002).


21

Klappmyss Formation; this formation belongs to the Sassendalen Group. This

formation shows a typical coarsening upward succession with medium to dark grey

shales at the base followed by interbedded shales, siltstones and sandstones. Its age

has been suggested to be Smithian to Spathian and it tends to thicken and fine

northwards from the Southern margins of the Hammerfest Basin. This unit was

deposited in a marginal to open marine environment with some renewed northwards

coastal progradation from the early Smithian transgression.

2.3.1 Trap

The major structural element cutting through the Goliat Field is the Troms-Finnmark Fault

Complex (TFFC), which is an old zone of weakness. This fault complex is characterized by

listric normal faults and associated hanging-wall roll-over anticlines and antithetic faults

(Faleide et al., 1984, Gabrielsen et al., 1990, Dore, 1995). This fault complex has been

reactivated several times up to Eocene times. The Goliat and Nucula oil discoveries are

located on roll over structures fairly close to the TFFC (Figure 2.10). As described by Ohm

and Karlsen (2008), most cap rocks located towards the basin flanks are usually thinner, more

faulted and contain more silt compared to the same cap rock close to the basin center. As a

result, the sealing capacity of cap rocks closer to the periphery of the basin, is generally

poorer than those found close to the basin axis. As a consequence, partly leaking cap rocks

most likely found in the periphery of the basin would most likely contain oil, after leaking off

accumulating gas. This is typical of Sales type II or III traps and may explain the

predominance of oil over gas in the Goliat field. There exists a correlation between the

expected hydrocarbon phase and the cap-rock quality in the traps (Ohm and Karlsen, 2008).


22

Fig. 2.10 Base cretaceous unconformity depth structure map. Seismic profile shows thinner

and more faulted cap rocks (Jurassic and Cretaceous), in the Goliat area than farther out in

the basin (modified from Ohm and Karlsen, 2008).

Fig. 2.11 Gamma ray log (API) for well 7122/7-3 showing some source rock intervals (circled

in red) including cap rock horizons. Upward coarsening and fining sequences are shown with

arrows.


23

The two main cap rocks of interest in this study are represented by the Fuglen Formation and

the base of the Snadd Formation as highlighted in Figure 2.11. The Fuglen cap rock is a

relatively tight cap rock and acts as a good seal, retaining a small gas cap in the Tubåen

Formation after uplift. On the other hand the deeper cap rock (base of the Snadd Formation),

has a relatively poor seal. Based on the gamma ray log, it is seen to be relatively coarser than

the Fuglen Formation and is representative of a Sales type II / III trap. Several fill and

subsequent spill scenarios may occur during Tertiary migration of hydrocarbons from existing

traps in an uplifted sedimentary basin. After uplift the pressure reduction will causes any

existing super critical petroleum phase (if present), to exsolve out some gas and form a two

phase fraction of oil and gas. Good seals (tight cap rock), have a higher probability to retain

gas as illustrated in Figure 2.12.

Fig. 2.12 Correlation between hydrocarbon phase and cap-rock quality (adapted from Ohm

and Karlsen, 2008).


24

2.4 Exploration challenges on the Barents shelf

There are several factors (e.g. geological and commercial) that need to be taken into

consideration when exploring for hydrocarbons on the Barents shelf. A number of these

factors summarized in this study have been described in Dore (1995). The Barents Sea is

relatively immature with respect to hydrocarbon exploration, when compared to areas like the

Northern North Sea. Some of these challenges include;

Cenozoic exhumation is the primary geologic consideration which renders

hydrocarbon exploration challenging on the Barents shelf. Residual columns of oil

have been found beneath several gas finds indicating that the structures were initially

or partly filled with oil in the past (Dore, 1995). In terms of the source rocks, cooling

as a result of uplift would have caused mature source intervals to stop generating more

hydrocarbons. With respect to reservoir rocks, they are of a lower quality than

expected at a given depth, because they are overconsolidated. Seal breaching and

resultant spilling of hydrocarbon accumulations from traps also occurred as a result of

uplift. The pressure release and subsequent expansion of gas adds the number of

variables involved in redistribution of the oil to distant traps which would have

otherwise not been charged (Ohm and Karlsen, 2008).

Other commercial considerations involved in hydrocarbon exploration in the Barents

shelf, have to do with the distance to potential markets. Harsh arctic climate,

remoteness of the area and the need for environmental precautions also adds

constraints to exploration in this area. Water depths here are averagely 300m deeper

than most oil fields in the North sea (Dore, 1995), and therefore need more

technologically advanced solutions for development ,such as subsea installations or

floating production systems. Oil finds are commercially more desirable. The

predominance of gas over oil needs additional processing facilities, to convert the gas

to liquefied natural gas (LNG) before being easily transported through long distance

pipelines. All these factors decrease the economic feasibility of hydrocarbon

exploration.

Chapter 3. Compaction and rock properties

25

CHAPTER 3 COMPACTION AND ROCK PROPERTIES

3.1 Introduction

Sedimentary basins usually go through a continuous cycle of creation and infill of

accommodation space. The infill usually takes place layer after layer, and results in a

stratified basin infill. The physical and chemical properties of these sediments change

significantly with increasing burial depth, from loose sediments to consolidated rocks. The

grain fabric undergoes an insignificant reduction in volume so that the increase in bulk

density with burial is linked to a reduction in pore volume. This reduction in the bulk

sediment volume (porosity loss), and resultant increase in bulk density as a function of burial

is generally referred to as compaction. Sediment compaction is driven towards higher density

(lower porosity) by mechanical compaction following the laws of rock and soil mechanics and

by chemical compaction controlled by thermodynamics and kinetics independent of the stress

(Bjørlykke and Jahren, 2010). Log derived rock properties such as velocity, porosity and bulk

density, change as a function of both mechanical (effective stress) and chemical compaction

(mineral thermodynamics and kinetics). These rock properties are closely related to each other

and change in a predictable pattern with increasing amount of compaction. Compaction

induced changes in these petrophysical rock properties, provide insight into the nature of, and

transition between mechanical and chemical compaction. The prediction of how the above

mentioned rock properties vary with depth is vital for input in seismic interpretation, depth

conversions, modeling and exhumation characterization (Marcussen et al., 2009).

This chapter will provide a brief theoretical framework as a base for subsequent discussions.

A description of the data methodology used to is also presented and the results subsequently

explained in the discussion.

3.2 Theoretical Background

3.2.1 Mechanical Compaction

Mechanical compaction starts immediately after sediment deposition, as the subsequent

overburden increases. In most sedimentary basins mechanical compaction due to increased

effective stress controls the changing rock petrophysical properties down to temperatures


26

between 70-800C. The effective stress (which represents the difference between the total

stress and the pore pressure), determines the extent of mechanical compaction. This effective

stress is transmitted through the grain framework and will be concentrated on the (largest)

grains, which experience a grain – to – grain contact stress higher than the overall average

effective stress. This eventually leads to grain reorientation and grain crushing, and the rock

becomes less compressible. Data from the shallow parts of the sedimentary basin dominated

by mechanical compaction show good agreement with experimental compaction data and can

be used to predict reservoir properties at depths corresponding to temperatures less than 70 –

80 0C, before the onset of chemical compaction (Marcussen et al., 2010).

Several factors influencing mechanical compaction such as mineralogy, sorting, grain size,

and shape (Figure 3.1), are closely linked to the provenance, duration and mechanism of

transport of the sediments into the sedimentary basin (Mondol et al., 2007, Mondol et al.,

2008, Mondol et al., 2009a). Poorly sorted sediments tend to compact readily at relatively

lower effective stresses than well sorted sediments, as the finer grained material can

adequately fill in the space between the coarse grains (Mondol et al., 2007). Rough grains on

the other end limit the amount of mechanical compaction due to grain rearrangement, when

compared to smooth, well rounded grains. Larger or coarse grained sediments of the same

mineralogy will however experience greater amount of grain crushing during mechanical

compaction as these grains feel a higher contact stress than fine grained sediments (Mondol et

al., 2009a). As a result of this, smectite rich mud which is fine grained and with a large

specific surface area, will compact much less readily than kaolinite rich mud which is coarser

and with a smaller specific surface area. The same observation is true for sandstones (Figure

3.2).


27

Fig. 3.1 Plots of petrophysical and acoustic properties of brine-saturated kaolinite

aggregates as a function of vertical effective stress (adapted from Mondol et al., 2008).

Fig. 3.2 Effect of sand grain size on mechanical compaction with increasing stress (adapted

from Bjørlykke and Jahren, 2010).

For the finer grained sediments, the effective stress due to increasing overburden is distributed

over larger number of grain contacts and the stress per grain contact is correspondingly lower.

The physical properties of mudstones in general will not depend only on geometrical

constraints but also on the composition of the pore fluids (Bjørlykke, 2010).


28

3.2.2 Chemical Compaction

Chemical compaction involves dissolution, precipitation and cementation of the grain

framework. Unlike mechanical compaction, chemical compaction is a function of

thermodynamics and kinetics (involving a time – temperature integral) during burial. Burial

depths corresponding to temperatures higher than 70 – 80 0C usually corresponds to the

transition between mechanical and chemical compaction for siliciclastic rocks. The exact

depth range at which this transition occurs varies from basin to basin depending on the

geothermal gradient.

Volcanic sediments and sediments containing fossils like diatoms and siliceous sponges tend

to be rich in Opal A (amorphous silica) and represent an important source for quartz cement

precipitation at a relatively shallower depth compared to the onset of quartz cementation from

pure sands and shales (Bjørlykke and Jahren, 2010). Dissolution of fossils rich in Opal A,

cause the pore water to be supersaturated with silica relative to quartz and could induce

precipitation of micro quartz on clastic quartz grains at relatively low temperatures. This

phenomenon is important in preserving porosity with depth. At higher temperatures pore

water is only slightly supersaturated with respect to quartz after minerals like smectite, Opal

A and Opal CT have reacted. This slight super saturation is not sufficient to precipitate quartz

cement on micro quartz coatings (Aase et al., 1996).

Re-crystallization of calcareous organisms and meteoric flushing of aragonite is an important

source for calcite cementation. Aragonite and high Mg calcite represent unstable carbonate

mineral phases. Fluvial sands with less biogenic production in fresh water have a lower

probability of calcite cement compared to shallow marine sandstones.

Very small amounts of cement will prevent grain rearrangement and significant grain

framework stiffening. This will result to an abrupt increase in the observed velocity (Storvoll

et al., 2005). The temperature range of 70 – 80 0C represents the activation energy required

for the transformation of thermodynamically unstable smectite to illite via mixed layer

minerals. A simplification of this reaction is shown below:

Smectite + K-feldspar = Illite + Silica + H2O


29

For this reaction to proceed, the silica released in the process needs to be precipitated as

quartz cement. A readily available k+

source, (such as from the dissolution of k-feldspars),

needs to be present.

The formation of stylolites (Figure 3.3) from pressure solution is critical for the destruction of

porosity in sandstones and grain framework stiffening (Bjørlykke and Jahren, 2010).

Mudstones may survive chemical compaction to greater depths in the absence of the more

thermodynamically unstable minerals like smectite.

Fig. 3.3 Quartz cement formation in sandstones and grain coatings (Byørlykke and Jahren,

2010).

At higher temperatures of about 130 0C and in the presence of K-feldspar, kaolinite becomes

unstable. It then reacts with K-feldspar, leading to the precipitation of illite and quartz cement.

The presence of K+ ions and removal of the silica from solution is necessary for the forward

reaction to be successful. Below is a simplified expression of this reaction:

KAlSi3O8 + Al2Si205(OH) 4 = KAl3Si3O10(OH) 2 + SiO2 +2H2O

K- Feldspar + Kaolinite = Illite + Quartz

The precipitation of quartz cement in either sand or clays has the effect of reducing the

porosity and increasing the grain–to-grain contact area. This makes the rock less sensitive to

changing effective stresses and inhibits further volumetric loss by mechanical compaction as a

result of stabilizing the grain framework. Very low geothermal gradients and fast rates of

deposition and subsidence in cold basins will enhance mechanical compaction and grain

crushing at even greater depths, unless an overpressure exists to counter the increasing

effective stress (Bjørlykke and Jahren, 2010). The volume of quartz cement that can be


30

precipitated is not only limited by the temperature – time integral but also by the total grain

surface available for quartz cement precipitation (Walderhaug, 1994).

Fig. 3.4 Grain coating by chlorite, well 6506/12-10, depth 5024.50m RKB, Smørbukk Field,

Haltenbanken (adapted from Byørlykke and Jahren, 2010).

The presence of grain coatings such as micro quartz, detrital clay, asphalt (bitumen) and

chlorite could preserve much of the primary porosity in deep seated reservoirs (Figure 3.4).

Quartz cementation will continue to cement the rock and further reduce the porosity even

during uplift (Bjørlykke and Jahren, 2010) as shown in Figure 3.5, provided the temperatures

stay above 70 – 80 0C which represents the minimum activation energy required for the

process. A simplification of the temperature controlled nature of this transition is shown in

Figure 3.5. When rocks have been subjected to higher effective stresses in their geologic past,

due to exhumation and glacial loading, they tend to have greater bulk and shear modulus than

corresponding rocks at the same present day burial depth. As a result they are referred to as

overconsolidated. Chemical compaction provides rocks with a substantial increase in bulk and

shear modulus as a result of grain framework stiffening. This phenomenon on the other hand

is referred to as pseudo- over consolidation, because the strength of the rock is not attributable

to higher prior effective stresses (Bjørlykke and Jahren, 2010).


31

Fig. 3.5 Diagenesis as a function of temperature and time (adapted from Byørlykke and

Jahren, 2010).

3.3 Materials and Methods

Overconsolidation of rocks due to uplift and Pseudo overconsolidation due to cementation of

the grain framework, gives reservoir rocks greater bulk and shear modulus than they would

normally have at the same depth, in the absence of these two factors. The rock properties

change and the reservoir quality becomes less than expected. This has consequences for

production. However delineating in advance at which target depth the reservoir properties are

hampered by cementation, will assist future development of a drilled prospect. Quantifying

the amount of uplift will give an idea of the degree of overconsolidation. The pressure release

also has consequences as to what extent long distance migration is effective in charging

distant traps in the periphery of the basin. The amount of uplift will also help ascertain if

‘source kitchen’ is located at depths, deep enough to constitute a live petroleum system.

The Vp, density, neutron logs were used to investigate changing rock properties due to

compaction as a function of burial depth. The gamma ray log, published regional data and

well completion reports have been used as the main lithologic control with depth for the

different petrophysical logs under investigation.

Three of the wells (7122/7-1, 7122/7-2, and 7122/7-5A), have a relatively poor and

incomplete sonic log coverage for the drilled formations and are not incorporated in further

compaction analysis. Two of the wells chosen for further analysis (7122/7-3 and 7122/7-4)

have proven hydrocarbon reserves and therefore the velocity-depth trend will be subject to


32

effects of varying pore-fluid saturation and overpressure effects. The reference well chosen

(7122/7-3) has the deepest depth coverage up to the Tempelfjorden Group also has direct

recordings of shear wave velocity and used for rock physics AVO modeling for selected

reservoir zones (Chapter 4). In all plots, the velocity is given in meters per second and the

depths are displayed in meters below sea floor (BSF).

A composite velocity versus depth plot comprising data from the 6 wells, was compared with

published data (used as a geophysical reference to investigate exhumation and possible

overpressure intervals); a first order linear velocity-depth trend based on Storvoll et al. (2005),

a kaolinite–silt curve from experimental laboratory compaction data based on Mondol 2011

(personal communication), and a Cenozoic marine shale velocity-depth trend line based on

Japsen (1999). These published trend lines are based mainly on shale data from several wells

in North Sea but can equally used to calibrate exhumation estimates from the Barents shelf

even though data from the Barents shelf will generally show a higher velocity-depth gradient

considering extensive Cenozoic exhumation and subsequent increases in geothermal gradients

and thermal exposure of sediments over time leading to a higher probability of extensive

quartz cementation.

The marine Cenozoic shale trend published by Japsen (1999), tends to agree more with shales

richer in smectite or illite. Meanwhile in estimating the linear trend based on Storvoll et al.

(2005), velocity data from carbonates and salts were not included. This linear trend shows a

close agreement to data published by Japsen (1999). The Storvoll et al. (2005) linear trend is

just a first order approximation, simply represented by equation 3.1;

Z =1.76Vp – 2600 (3.1)

Where Z = depth (meters) and Vp = P-wave velocity (meters per second).

The experimental compaction trend of brine-saturated kaolinite–silt (50:50) mixtures was

used as a base to estimate the amount of exhumation in the Goliat Field. The choice of the

laboratory kaolinite-silt compaction trend rather than end members sand, silt or clay is based

on the fact that, mudrock in nature occurs mostly as a composite mixture of sands, silts and

clays of varying weight fractions. The shallowest formations (such as Kolmule, Kolje and

Knur) in the wells under investigation are rich in shale and silt beds, still in the mechanical

compaction regime.


33

No analytical work on textural and mineralogical compositions has been carried out on the

formations mentioned in this study, and the trends described in the well log data are based on

published information about group and formation lithology from NPD (Norwegian Petroleum

Directorate) website and Worsley (2008). Problems related to depth measurements in this

dataset are minimized because all the wells are vertical (except well 7122/7-5A which is a

side track from well 7122/7-5) and therefore assumed to correspond closely to the true

vertical depth. A more precise variation of velocity with depth is presented in this study as the

entire log data sets from six wells have been used without any statistical averaging. This

results in the wide data spread observed as a consequence of varying lithology and degree of

diagenetic alteration (subsurface heterogeneity).

3.4 Results

3.4.1 General porosity/density/Vp versus depth trends

There is an inherent interdependent relationship between the different log petrophysical

properties such as Vp, density and porosity as function of burial depth due to the combined

effects of mechanical and chemical compaction. For a given uniform lithology and constant

pore fluid and pressure, the density and Vp will increase as a function of burial depth,

meanwhile the porosity will correspondingly decrease.

The Vp-depth trends in all the wells generally tend to increase with depth, with exception of

some anomalous zones due to varying lithology, pore fluid and possible effects of pore

pressure (Figures 3.6 and 3.7). The gamma ray log used here serves as a lithologic control to

constrain the variation in Vp, porosity and bulk density with depth. Very high Vp values of

approximately 3500 m/s at shallow burial depths of 250 m (BSF) are observed in all the wells

except in well 7122/7-1.

It is clear from Figure 3.6 and 3.7 that the transition from mechanical to chemical compaction

for silicilastics (excluding the carbonates) tends to occur at similar depths in all the wells, at

approximately 600m. The transition defined here is based on the Vp-depth trend, at 600m

BSF (extrapolated across the other depth trends) where for the same lithology there is a rapid

increase in the Vp response. When correlated with the gamma ray log, it is confirmed that the

increase in Vp is not due to a change in lithology. The trend line representing chemical

compaction generally shows a smaller change in velocity with depth.


34

Fig. 3.6 Compaction trends observed in wells 7122/7-1, 7122/7-2 and 7122/7-3.


35

Fig. 3.7 Compaction trends observed in wells 7122/7-4, 7122/7-5 and 7122/7-5A.


36

Beneath the transition from mechanical to chemical compaction, there is a decrease in

observed Vp with depth in all the wells, and correlates with low bulk densities and high

gamma readings at the same depth interval. This corresponds to the major source rock in this

area (Hekkingen Formation). Both the neutron and density porosities tend to decrease with

depth. The density porosity was calculated using a constant matrix density of 2.65 g/cc which

is the matrix density of sandstone. There is therefore a deviation between both porosities in

depth intervals with high gamma. Negative density porosities are not shown as they

physically represent data points with matrix densities greater than 2.65 g/cc. There is an

apparent inverse relationship between the gamma and Vp depth trends in all the wells.

Generally the density shows a fairly linear increase with depth. Despite the general increase of

density with depth, there are zones showing a decrease in density with depth and these areas

correlate with high gamma values such as, at depths of about 700m BSF (Figures 3.6 and 3.7).

It is observed that depths with significant changes in the gamma ray log, show corresponding

changes in the density, Vp, and porosity depth trends due to a lithologic control on these

parameters.

3.4.2 Vp-depth trend for well 7122/7-3

Well 7122/7-3 is chosen as the reference well. This well has the deepest depth coverage, most

complete log suits and also direct measurements of shear wave velocities (Vs). The Vp-depth

trend in Figure 3.8 shows a general increase in Vp with depth. There is however a break in the

Velocity depth gradient with depth at about 600m BSF and is taken to be the transition

between the mechanical and chemical compaction for siliciclastic rocks in this well. Two

simplified trends (shown as solid red and green lines) can be made for mechanical and

chemical compaction as shown in Figure 3.8. The first trend corresponding to mechanical

compaction (red) from approximately 150m to 600m BSF shows a lower gradient of 0.85;

meanwhile the chemical compaction trend (blue) from approximately 650 m to 2350m has a

higher gradient of 1.68 compared to the trend above.

Velocities in the shallowest part of the trend belonging to mechanical compaction are less

than 2500 m/s. There is however an exception in zone A. The Zone A located at shallow

depths of about 250m BSF shows an abrupt increase in the Vp-depth trend up to 3500 m/s.

Zone B, C, and D show a decrease in Vp with depth and deviate from the general trend line

(green) for chemical compaction.


37

Fig. 3.8 Well 7122/7-3 Vp-depth trend and anomalous zones.

3.4.3 Transition from mechanical to chemical compaction

The transition from mechanical to chemical compaction has been defined based on the Vp-

depth trend and a shear modulus-porosity cross plot (Figure 3.9). This transition has been

defined based on the sudden sharp increase in Vp for a given lithology at present day depths

of approximately 600m below sea floor. The depth at which this transition occurs is seen to

occur at similar depths in all the wells as defined in Figure 3.6 and Figure 3.7. The depth at

which this transition occurs as defined from the Vp depth trend correlates with the result

obtained from the shear modulus/porosity cross plot in Figure 3.9.


38

Fig. 3.9 Shear modulus-Porosity cross plot color coded with Vshale and depth.

The shear modulus has been computed using the directly measured shear velocity and bulk

density. A cross plot between shear modulus and density porosity (calculated using matrix

density of 2.57 g/cc for shales only, in well 7122/7-3 shows a ‘knick’ point transition from

mechanical to chemical compaction. Careful analysis indicates that chemical compaction sets

in with approximately 15% porosity left in the shales and at a shear modulus of approximately

5 MPa. The average shear modulus for data points within the mechanical compaction regime

is about 4 MPa with a wider range in porosities up to 35%. The entire data set shows a wide

range of shear modulus from approximately 3-17 MPa. There is a significant increase in the

shear modulus after onset of chemical compaction, and a continuous reduction in porosity.

The shear modulus-porosity plot color coded with depth in Figure 3.9 shows the transition

from mechanical to chemical compaction (‘knick’ point) at about 613 m BSF. This closely

agrees with the transition obtained using the Vp-depth trend of 600m BSF.

The bottom hole temperature in well 7122/7-3 is 73°C at a true vertical depth of 2725m (NPD

Factpages). The sea floor temperatures, based on Coastal Water and North Atlantic Water is

>2°C and 3°C respectively (Loeng, 1991). An estimated geothermal gradient can then be

made of 29.20C/km. This gradient is slightly lower than the actual, because the bottom hole

temperatures recorded are less than the true temperatures as the drilling mud tends to cool the

down hole assembly during drilling. However geothermal gradients measured from a shallow

well on Spitsbergen-banken are slightly higher at 31°C/km (Solheim and Elverhoi, 1993).


39

The transition from mechanical to chemical occurs at approximately 1.3 km (after applying a

700m correction for uplift in Figure 3.12). This depth corresponds approximately to

temperatures of 41°C using a geothermal gradient of 31°C/km.

3.3.4 Uplift estimation

Only sorted shale data has been used for uplift estimation. Figure 3.10 shows a composite plot

comprising of shales only from selected wells with sufficient coverage. When compared with

experimental compaction curves of Kaolinite – silt (50:50) and the pure Kaolinite trends,

there is a significant mismatch between the published trends and the composite Vp-depth

profile. This mismatch is seen even for the mechanical compaction regime. This deviation is

greater for the chemical compaction regime. An even greater deviation is seen for the trend

lines proposed by Japsen (1999) and Storvoll et al. (2005). This observation could be

interpreted as the overconsolidation of the shales in this area due to uplift. It is therefore

important to quantify this exhumation (uplift and erosion) as it has consequences on the

petroleum system in the Goliat field.

The data presented has been sorted based the volume of shale (Vshale>80%). The composite

shale trend shown in Figure 3.10 displays mainly data points corresponding to well 7122/7-3

in the mechanical compaction domain, and as a result further exhumation estimates are based

on this well only.

Fig. 3.10 Composite shale trend compared with clay –clay and clay –silt curves.


40

Fig. 3.11 Exhumation estimates using clay –clay and clay –silt curves.

Different estimates for exhumation were obtained by extrapolating data points (well above the

transition into chemical compaction) onto the different experimental laboratory compaction

curves as shown in Figure 3.11. The largest exhumation estimates of 1500m were observed

using the pure kaolinite curve, meanwhile the kaolinite – silt curve resulted to a minimum

estimates of 700m, which is almost half that obtained from the pure kaolinite curve. The

kaolinite – smectite estimates are much closer to the kaolinite estimates.

The Vp-depth profile for well 7122/7-3 shown in Figure 3.12 (after correction for exhumation

based on the kaolinite – silt compaction curve in Figure 3.11), illustrates that, the part of the

curve belonging to mechanical compaction (blue) shows a better fit to the Kaolinite – silt

trend compared to the first order linear velocity trend and the Cenozoic marine shale trend.

The shale trend by (Japsen, 1999) shows a relatively close fit only to shallowest Torsk

Formation of the Sotbakken Group deposited during the opening of the Norwegian –

Greenland sea.


41

Fig. 3.12 Well 7122/7-3 Vp-depth trend before and after exhumation showing transition from

mechanical to chemical compaction.

Fig. 3.13 Complete composite well data before and after exhumation.


42

The composite Vp-depth plot for all six wells in Figure 3.13 shows a closer match to

experimental compaction clay data after applying 700m correction for uplift derived from the

kaolinite – silt (50:50) curve.

3.4.5 Sand and shale compaction trends

The sand and shale trends for well 7122/7-3 have been sorted out based on Vshale, computed

from the gamma ray log in the same well. Both sand and shale trends shown in Figure 3.14

are in the zone corresponding to chemical compaction. The blue straight line connector

represents the mechanical compaction regime. Data points within the mechanical compaction

regime are composed dominantly of silts, based on Vshale color code. The shale trend

illustrated shows a smaller change of velocity with depth compared to both sand trends. The

grey interval represents zone C defined in Figure 3.8.

Fig. 3.14 Variations in sand and shale compaction trends.


43

Zone C shows a lower shale velocity anomalous to the general shale trend line in red. Both

sand trends have similar gradients and are separated by the thick shale unit with a lower

velocity in Zone C. The beginnings of both sand trends correspond to the Tubåen and Kobbe

reservoir units respectively.

Fig. 3.15 Shale Vp/bulk density/porosity-depth trends.

The data points shown in Figure 3.15 represent shale data points only in well 7122/7-3

selcted based on Vshale > 80%. A bulk density of 2.57 g/cc was then used to compute the

density porosity (DPHI). The density and Vp tend to increase with depth meanwhile the

density porosity shows an inverse relationship. All three parameters are intricatley linked and

predictably vary with depth as a function of burial diagensis. Data points from all three Vp/

bulk density/ porosity depth trends above the transition from mechanical to chemical

compaction, deviate from experimental compaction kaolinite – silt (50:50) curve (Mondol

2011, personal communication). This is due to exhumation, as demonstrated in the


44

corresponding plots beneath showing a much closer fit to experimental laboratory compaction

data after applying a correction of 700m upflit to the data.

3.3.6 Effect of pore fluid and pore pressure

Most of the shales represented in Figure 3.16 are already in the chemical compaction regime.

The Kobbe oil reservoir is shown in the grey circled zone at approximately 2600m BSF after

correcting for exhumation. This demonstrates clearly the effect of varying pore fluid type on

the Vp-depth profile in well 7122/7-3. The shale trend shown in red in well 7122/7-3 shows

clearly an anormalously low Vp. The grey straight line connector is an approximate best fit

line for the shale trend. Meanwhile the circled grey zone represents a cluster of shales which

show possible overpressure effects. This circled cluster of shales is the same as zone C in

Figure 3.8.

Fig. 3.16 Well 7122/7-3 showing possible overpressure effects.

The smectite curve shows the lower bound while the kaolinite – silt curve represents the

upper bound for the experimental data. The shales deviate to higher Vp values away from


45

the upper bound. The shale trend shows a bettter match to Japsen (1999) and Storvoll et al.

(2005), compared to the experimental compaction trends.

3.3.7 Effect of source rock on Vp-depth trend

The major source rock in this area which is also widely distributed throughout the entire

Barents Shelf is represented by the marine dominated shales of the Upper Jurassic – Lower

Cretaceous Hekkingen Formation. The high kerogen content (Figure. 3.17) is reflected in the

high gamma log readings, and low bulk density compared to shales with little or no organic

content. The Hekkingen Formation is seen to display a coarsening upward succession with the

base of this unit having a significantly higher gamma reading and possibly more organic

content. The velocity inversion seen in this interval is also present at similar depths in other

wells.

Fig. 3.17 Gamma, Vp, deep resistivity and bulk density petrophysical logs for the source rock

interval (Hekkingen formation).


46

Well 7122/7-5 was selected as the best candidate for the effects of source rocks on Vp,

because it is a dry well with no hydrocarbons and only brine. As a result there is no influence

from varying pore fluids in the underlying Kapp Toscana Group. The velocity inversion

(velocity decreasing with depth) shown in red in Figure 3.18, starts soon after the onset of

chemical compaction as shown by the transition with the orange line at present day depths of

about 610m BSF up to about 750m BSF. As such Vp-depth variations in this well are

assumed to be as a result of varying lithology. The corresponding logs for this interval show

very high gamma readings of up to 250 API, and relatively low resistivity less than 75 ohm-

m. The Hekkingen Formation is sandwiched between the Fuglen and Knurr Formations and

has a bulk density lower than these two, partly due to its high organic content, as shown in the

plot in Figure 3.17. Despite the low velocities present in the source rock, it is however still

higher than overlying shales of the Kolmule, Kolje and Knurr Formations.

Fig. 3.18 Source rock velocity inversion.


47

3.4 Discussion

3.4.1 Relationship between porosity/ density/ Vp versus depth trends

There are significant deviations of the composite well data from the Goliat Field compared to

the published data. A number of factors can be responsible for this deviation, but separating

the independent effects of these factors require of combination of other petrophysical logs

such as the density, resistivity and gamma ray curves.

Some of these factors include;

Lithology variations

Pore fluid and pore pressure effects

Transition from mechanical to chemical compaction

Uplift and erosion (exhumation)

Generally the P-wave velocity (Vp) tends to increase with depth due to corresponding

increases in the bulk and shear modulus with depth. The velocity is also inversely

proportional to density. Despite the general increase in density with depth as a function of

burial and compaction, the velocity still increases because the increase in bulk and shear

modulus with depth is greater than the increase of density with depth.

Compaction tends to increase the stiffness, and reduce the porosity (Mondol et al., 2007) in

the rock framework through combined effects of increasing effective stress (mechanical

compaction) and through dissolution of less stable mineral phases and precipitation of

thermodynamically more stable ones (Bjørlykke and Jahren, 2010). The main controls for

Velocity however are the porosity and the microfabric (Fawad et al., 2010). The Shallow

carbonates (NPD Factpages) stand out with very high Vp of approximately 3500 m/s at very

shallow burial depths of 250m BSF. Aragonite and high Mg calcite are unstable carbonate

mineral phases and react at much lower temperature and depths compared to common

minerals found in siliciclastic rocks. These wells are relatively closely spaced to each other

and are therefore expected to have similar geothermal gradients. The trend line representing

chemical compaction generally shows a smaller change in Vp with depth, possibly due to

increased strength of the rocks due to quartz cementation.


48

The variation in Vp with depth will further be explained using well 7122/7-3 as a reference as

it has the most complete data coverage.

Fig. 3.19 7122/7-3 anomalous zones and corresponding petrophysical logs.

From Figure 3.19, four different zones A, B, C, and D can be can be distinguished, which

deviate from the two trend lines for mechanical and chemical compaction.

Zone A: belongs to the Kviting Formation of the Nygrunnen Group. This late Cretaceous unit

was formed during periods of platform uplift, and it’s rich in claystone with limestone

interbeds and calcareous sand units (Worsley, 2008). The carbonates present in this formation

are responsible for the very high velocities of up to 3500 m/s at shallow depths of 250 m BSF.

The activation energy for calcite cementation is much lower and occurs at much shallower

depths unlike quartz cementation (Bjørlykke and Jahren, 2010).

Zone B: corresponds to source rocks of the Upper Jurassic – Lower Cretaceous Hekkingen

and Fuglen Formations of the Adventdalen Group and part of the Kapp Toscana Group

belonging to the Tubåen Formation (reservoir). The overlying source rocks are rich in marine

dominated kerogen shales of up to approximately 20% TOC (Worsley, 2008). These were

deposited during a renewed transgressive cycle with reduced coarse clastic input into the

basin. The high kerogen content is reflected in the high gamma ray log readings, and low bulk


49

density, compared to shales with little or no organic content. The velocity inversion seen in

this log interval is also present at similar depths in other wells.

The effect of source rocks on Vp will be discussed subsequently. Despite the low velocities

present in the source rocks in Zone B, it is however still higher than overlying shales of the

Kolmule, Kolje and Knurr Formations. These formations were deposited when there was a

reduction in sea level, with better circulation of bottom currents except around local highs in

the basin, thus establishing oxic conditions and less preservation of kerogen. The higher

velocity of the source rock with more kerogen compared to the overlying shales is due to the

transition from mechanically dominated compaction to chemical compaction at approximately

the same depth where there is a lithologic transition. The Fuglen Formation acts as cap rock

for the Kapp Toscana reservoir sands beneath. The combined effects of the source rock and

Tubåen gas cap (NPD Factpages) possibly reduce the Vp in this interval.

Zone C: shows very low velocities in the chemical compaction domain. This interval shows

low gamma API in the Kobbe sands then a fining upward succession at the base of the Snadd

formation, inferred from gradual increasing gamma. This zone shows a transition from high to

low gamma values and corresponds to the transition between the transgressive shales at the

base of the Snadd formation to the sands of the Kobbe reservoir. The base of the Snadd

formation has thick marine shale units deposited during a regional transgression during which

most of the structural highs were submerged. During this time in the Carnian age, there were

very high rates of subsidence on the Barents shelf (Worsley, 2008). The overlying sandstone

units above Zone C show significantly higher velocities. These Carnian sandstones were

sourced from the Baltic Shield and tend to be texturally and mineralogically mature. These

sands would have had a relatively good primary reservoir quality at shallow depths. However

at burial depths corresponding to chemical compaction, the mature sands will then undergo

pervasive quartz cementation possibly sourced from stylolites (Bjørlykke and Jahren, 2010).

This could possibly explain the high Vp values for these Carnian sands above Zone C.

Zone D: represents part of the Sassendalen Group deposited at a time when subsidence rates

were also high, in the Late Triassic. This zone shows interbedded shale source rock intervals

deposited under anoxic conditions, which show up with high gamma, low resistivity and low

bulk densities as seen from the corresponding logs. These source rocks are time equivalent to


50

the oil shale Botnehia formation on Svalbard with approximately 10 %TOC (Worsley, 2008).

These intervals account for the low velocities observed in this trend.

3.4.2: Uplift Estimation

Cores from shallow well sections are not usually available. As a result laboratory compaction

of synthetic mudstones (different clay fractions) with a control on the initial mineralogy prior

to compaction, (Fawad et al., 2010) can be employed as a model to compare data confidently

known to belong to the mechanical compaction regime. In nature pure clay fractions are

seldom found, and therefore in this study different mixtures of clay – clay and clay – silt have

been used to quantify the amount of exhumation. The term ‘’silt ‘’ used here mineralogically

refers to quartz while ‘’clay’’ refers to the phyllosilicate minerals kaolinite and smectite, as

used in Fawad et al. (2010), except otherwise stated. The clay particle sizes for the

experimental compaction curves used in this work are based on Mondol et al. (2008a), and

range between 0.4µm-30µm while the silt used is between 4µm-40µm.

The composite shale trend for three wells with significant coverage (Figure 3.10), was sorted

based the volume of shale (Vshale>80%). This composite shale plot displays mainly data

points corresponding to well 7122/7-3 in the mechanical compaction domain, and as a result

further exhumation estimates are based on this well only. Monomineralic shales seldom exist

in nature, and usually occur with some silt. Therefore, the estimates derived from the

Kaolinite–silt curve have been preferentially considered in this study as opposed to estimates

from the pure kaolinite and kaolinite–smectite curves. Shale data points in the mechanical

compaction domain, which are greater than the kaolinite–silt (50:50) curve possibly represent

samples with kaolinite–silt ratios with more clay than silt, (matrix supported) resulting in

enhanced grain reorientation and stiffening of the clay system (Fawad et al., 2010).

Exhumation estimates of 700m imply that these shales have been exposed to higher pressures

of about 7MPa in the past prior to exhumation, leaving the shales over consolidated. Detailed

core analysis in future will be required to prove this relationship. However the kaolinite–

smectite curve provides just a minimum exhumation estimate (700m) for the Goliat field.

From this study it is not possible to give an exact exhumation estimate due to lack of proper

mineralogical control.


51

Fig. 3.20 Tentative uplift map based on Vitrinite reflectance data (modified from Ohm and

Karlsen, 2008).

However the amount of uplift should range between the upper limit (1500m) defined by pure

kaolinite to a lower limit (700m) defined by the mixed kaolinite-smectite curve. The upper

limit (1500m) defined for exhumation in this work closely agrees with the upper limit defined

by vitrinite reflectance data in Figure 3.20 from a suite of different wells on the Barents shelf

(Ohm and Karlsen, 2008).

The estimates provided in this study represent the possible cumulative uplift in this area,

without regard to the different episodes over which it occurred. A summary subsidence curve

(Figure 3.21) for different areas on the Barents shelf including the Hammerfest basin has been

presented by Ohm and Karlsen (2008).

However an intermediate estimate of 1000m may be a more likely scenario. It can be seen

from Figure 3.20 that the amount of uplift decreases westward toward the Tromsø Basin and

increases north of the Hammerfest basin (circled in red) towards Stappen High.


52

Fig. 3.21 Subsidence curves for different regions on the Barents shelf (adapted from Ohm and

Karlsen, 2008).

Detailed XRD and petrographic SEM studies for the mudstones in the zone of mechanical

compaction will provide some control on the mineralogy. This will provide the appropriate

silt-clay ratio to the used for exhumation estimates and possibly constrain the upper limit for

exhumation in the Goliat Field.

3.4.3 Transition from mechanical to chemical compaction

For a normally subsiding sedimentary basin, with average geothermal gradients (350C/Km)

the temperature controlled transition between mechanical and chemical compaction is

expected to occur at approximately 2000m BSF. The shallower transition (600m) observed in

all the drilled wells in the Goliat Field provide a clear indication of the effect of uplift.

The shale trend by Japsen (1999) in Figure 3.12 shows a relatively close fit only to the

shallowest Torsk Formation of the Sotbakken Group deposited during the opening of the

Norwegian – Greenland Sea. The associated volcanism and tuff generation was an important

source for smectite in this claystone dominated formation. This may explain the closer

relationship to the Japsen (1999) shale trend for this unit considering that this trend generally

shows a closer match to formations richer in illite or smectite. Of all the trends the sand trend

fits best with the mechanically compacted part of section, despite the fact that these shallow

formations are composed mainly of mudstones. The relatively high velocities for these


53

mudstones emphasize the fact that they are highly over-consolidated and have experienced

higher effective stresses in their burial history than is observed today.

Stiffening of the grain framework in the chemical compaction domain may possibly be

responsible for the large deviation between the experimental compaction of pure clay, clay –

clay and clay – silt mixtures. After correcting for exhumation using the kaolinite – silt

experimental curve, the depth at which the transition from mechanical to chemical

compaction occurs is approximately 1300m BSF (Figure 3.12). This transition corresponds

approximately to temperatures of 410C assuming a geothermal gradient of 31

0C/km. This is

lower than the expected 70 – 800C required for onset of quartz cementation in shales as a

result of illitization of smectite in the presence of K-feldspar (Thyberg et al., 2010).

A possible candidate responsible for cementation at this low temperature is the silica phase

transformation from Opal-A to Opal CT which may cause an abrupt increase in Vp due to the

precipitation of a critical volume of Opal CT which possibly strengthens the grain framework.

Lithological changes within the Opal CT field or conversion from Opal CT to quartz may also

be responsible. However evidence for the actual mineralogical transition responsible for the

log transitions will need to be confirmed in future XRD and SEM petrographic techniques.

Opal A is known to occur over a wide area in the Cenozoic of the Barents shelf (Roaldset and

He, 1995). Roaldset and He (1995) concluded that the transition from Opal-A to Opal CT

occurs at 1.4Km BSF meanwhile Opal CT to quartz transition occurs at 1.7 Km (Figure 3.22)

corresponding to temperatures of 40-450C and 50-55

0C respectively for both transitions. The

depth interval between both transitions is 300m. The expected thermodynamic temperature

for the transition between Opal CT and quartz is 60-650C. This transition temperature is above

the temperature (410C) observed in the Goliat Field.

The transition depths/ temperatures found for the opal A to opal CT and opal CT to quartz

transitions in well 7117/9-1 indicates approximately 300m of uplift in this well. If the

transition from mechanical to chemical compaction in the Goliat area is related to stiffening

within the opal-CT stability field, then based on well 7117/9-1 an estimate of 1100m of uplift

may be given for the Goliat area. If the transition between Opal CT and quart is responsible,

an additional 300m of uplift is inferred giving a total uplift of 1400m in the Goliat area. Both

estimates are based on similar geothermal gradients in well 7117/9-1 and the Goliat area.


54

Fig. 3.22 Mineralogy of well 7117/9-1(Roaldset and He, 1995).

The present day depth at which this transition occurs will vary across the Barents shelf due to

differential amounts of uplift. From the tentative uplift map shown in Figure 3.20, the amount

of uplift reduces west from the Goliat field. The control well (7117/9-1) with the XRD data is

located farther west from the Goliat, and as expected the present day depth of transition from

Opal A to Opal CT occurs (800m) deeper than in the Goliat. This is due to the larger amount

of exhumation experienced in the Goliat. Compared to well 7117/9-1, an additional uplift of

the Goliat Field between 800-1100m would be expected, depending on similar

paleogeothermal gradients in the two areas. Together with the 300m uplift found in well

7117/9-1 based on the present day Opal CT to quartz transition temperature, an estimate of

1100-1400m uplift in the Goliat area is expected.

In Figure 3.8, the trend corresponding to mechanical compaction (red) shows a gradient of

0.85, meanwhile the chemical compaction trend (blue) has a higher gradient of 1.68. The

physical implication of this relates a higher gradient (steeper trend) to a lower change in

velocity with depth. As such, in the chemical compaction domain, quartz cementation leads to

significant increase in strength and grain framework stability. Thermodynamically more

stable mineral assemblages are also expected to be formed. A combined effect of these factors

will lower the rock response to further compaction, and therefore a smaller change in Vp with

depth. Even a small amount of cement may inhibit compaction due to grain rearrangement

and crushing (Bjørlykke and Jahren, 2010).


55

The cross plot between shear modulus and density porosity (Figure 3.9) for shales only, in

well 7122/7-3 shows a ‘knick’ point transition at approximately 5 MPa. The significant

increase in the shear modulus after onset of chemical compaction up to about 17 MPa, may be

attributed to the fact that, quartz cementation will prevent further volumetric loss by grain

rearrangement. Increasing the effective stress increases the degree of clay mineral alignment

in shales and will yield a relatively low shear modulus in the mechanical compaction regime

(Fawad et al., 2010). The shear modulus is a good frame indicator because it is insensitive to

changes in pore fluid. The bulk modulus and Vp on the other hand are influenced by pore

fluid type and saturation. The transition depth of 613m BSF closely agrees with the value

obtained from the Vp-depth trend. A combined interpretation from both methods results to a

better constraint on the transition from mechanical to chemical compaction, though further

petrographic analysis from cores in this depth interval will be important to ascertain this

claim.

3.4.4 Variations in the sand and shale compaction trends

Sands and shales tend to compact at varying degrees and rates due to differences in grain size

and mineralogy. The shale trend illustrated in Figure 3.14 shows a smaller change of Vp with

depth compared to both sand trends. This implies that the shales are stiffer and more

consolidated than the sands at the same depth in the chemical compaction regime. As a result

they respond less to increasing compaction with depth. Sands have a larger grain size and tend

to compact more due to mechanical compaction by grain crushing, as a result of smaller

number and area of grain contacts, making the effective stresses high.

Mudstones are usually deposited with a higher depositional porosity than sandstones. At

greater burial depths, before the onset of chemical compaction, the inter-granular volume

(IGV) for clean sands is usually more than for mudstone. When quartz cementation starts, the

greatest effect may be most likely felt in the case with a smaller IGV, and the rock quickly

becomes completely cemented and responds less with increasing compaction (smaller change

in velocity with depth).


56

3.4.4 Effect of pore fluid and pore pressure

The dominant fluid in sedimentary basins is brine. Given normal sedimentation rates with

equilibrium compaction, sediments will undergo dewatering with depth. The probability of

generating an overpressure for the same fluid content is greater for finer grained sediments

with lower permeability. The deposition of shale under high sedimentation rates will result to

a situation of undrained compaction, leaving these shales at depth with an overpressure. Pore

pressures greater than hydrostatic will slow down or even prevent volumetric loss with depth

due to less increase of effective stress. This will lead to a larger inter granular volume before

the onset of chemical compaction and therefore will require much more cement to fill all pore

spaces. Therefore, overpressure during mechanical compaction represents an important

mechanism in preserving porosity with depth.

Most of the shales represented here are already in the chemical compaction regime. During

chemical compaction , temperature becomes the dominant controlling factor and effective

stress (and hence pore pressure) will have little controle on the porosity destribution with

depth (Bjørlykke and Jahren, 2010). As shown in Figure 3.16, this zone corresponds to the

Kobbe oil reservoir and shows high resistivities. This demonstrates clearly the effect of

varying pore fluid type on the Vp-depth profile in well 7122/7-3. The main controlling factors

are the bulk modulus and the density of the saturating fluids. The shear modulus is relatively

insensitive to changes in the pore fluid, but the compressiblity of hydrocarbon fluids is higher

than that of formation brine. As a result even with the lower oil density compared to brine,

there is still a significant reduction in Vp in that interval. The circled grey zone represents a

cluster of shales which show possible overpressure effects and corresponds to zone C in

Figure 3.8. This zone is the base of the Snadd Formation and is possibly an over pressured

interval. During this time in the Carnian age, there were very high rates of subsidence in the

Barents shelf (Worsley, 2008). Deposition of mudstones under high sedimentation rates may

result to undrained compaction with depth, and then develop an overpressure. The base of the

Snadd Formation acts as the cap rock for the Kobbe reservoir sands. Possible leakage of

hydrocarbons from the Kobbe reservoir into the Snadd cap rock may also account for the

decrease in Vp. However overpressure has little effect in the chemical compaction domain.

This is because chemical compaction in thermodynamically and kinetically controlled rather

than by effective stress. A fluid effect may therefore be the most likely cause for the strong

reduction in Vp.


57

The shales deviate to higher Vp away from the upper bound as shown in Figure 3.16 because

they have been over consolidated due to cementation.As a result they deviate from

experimental laboratory compaction data and the published trends.

3.3.5 Effect of source rock

The major effect of the presence of source rocks on the Vp-depth profile is to cause a

reduction in Vp with increasing depth. This velocity inversion with depth is shown in the

Hekkingen Formation in Figure 3.18. The Hekkingen Formation is one of the major source

rocks on the Barents shelf and is rich in marine dominated kerogen shales of up to

approximately 20% TOC (Worsley, 2008). These were deposited during a renewed

transgressive cycle with reduced coarse clastic input into the basin.

Despite the low velocities present in the Hekkingen source rock, it is however still higher than

overlying shales of the Kolmule, Kolje and Knurr Formations. These formations were

deposited when there was a reduction in sea level, with better circulation of bottom currents

except around local highs in the basin, thus establishing oxic conditions and less preservation

of kerogen (NPD Factpages). The higher velocity of the source rock with more kerogen

compared to the overlying shales in case is due to the transition from mechanically dominated

compaction to chemical compaction at approximately same depth where there is a lithologic

transition. A combination of several factors may explain the reduction in Vp with depth for

source rocks such as the Hekkingen formation in the Goliat Field. Source rocks usually have a

3D net work of compressible kerogen occurring mainly as laminae, with a preferred

orientation parallel to bedding. This induces anisotropy in velocity measurements, such that

vertical velocity recordings perpendicular to bedding are significantly lower (Stainforth and

Reinders, 1990). This anisotropy is enhanced in mature source rocks (such as the Hekkingen

Formation) during generation and expulsion of hydrocarbons as a result of adsorbed

hydrocarbons on the insoluble kerogen. Conversion of solid kerogen to hydrocarbons in

source rocks also increases the porosity. More important, is the geometry of the pores created.

Low aspect pores parallel to bedding with low incompressibility as opposed to capillary tube

type pore geometry may also further reduce Vp. Collapse of overburden load with increasing

vertical effective stress due to sediment loading may be responsible for inducing micro

fractures in source rocks parallel to bedding (Vernik and Liu, 1997). A combination of these

factors may explain the low Vp observed in the Hekkingen Formation

Chapter 4. AVO modeling

58

CHAPTER 4 AVO MODELING

4.1 Introduction

AVO (Amplitude Versus Offset) studies variations in reflection amplitudes with changing

distance (angle of incidence) between shot point and receiver for a given target reflector. Over

the years this technique has been used as a direct hydrocarbon indicator (DHI), when

conducting a bright spot analysis. Unconditional use of this technique without efficient pre–

processing, has led to several cases of both false positive and false negative indications of in

situ hydrocarbons in the subsurface. Unfortunately, bright spot anomalies are not just fluid

related but also lithology dependent. In order to condition the AVO analysis from a seismic

section, it is important to correct for spherical divergence, ensure removal of coherent noise

such as multiples and a normal move out correction (NMO) (Gelius and Johansen, 2010). In

addition to these routine corrections, the effects of variation in the overburden should be

accounted for, such as anisotropy effects in the case of a highly layered cap rock and

differences in mineralogy and transmission losses (Stainforth and Reinders, 1990). If these

corrections are not made, then the amplitude variations with offset on the subsurface target

may simply be due to variations in the overburden. Diagenetic alterations have also been

documented, (such as the conversion of smectite to illite, involving the precipitation of the

released silica as quartz) to be strong enough to be recognized as a seismic event (Thyberg et

al., 2010). The impedance contrast between oil and water in most non – biodegraded

conventional reservoirs is usually very small, and even smaller in situations where the

hydrocarbons in the subsurface are present in a pseudo – critical point where no distinction

exists between oil and gas (Batzle et al., 2005).

AVO modeling is a forward modeling approach that links the petrophysical reservoir

parameters to seismic elastic rock properties. There is a link between changing reservoir

parameters and the resultant seismic expression. For example, increasing the volume of shale

in the reservoir will increase the water saturation and reduce the effective porosity. A

combined effect of this has consequences for the resultant seismic response as the effective

reservoir parameters change. The link between these petrophysical log properties and the

effective rock seismic properties can be established using Gassmann’s equation (Gelius and

Johansen, 2010). Petrophysical logs contain significantly higher frequency measurements and

hence better resolution compared to seismic data with attenuation wave propagation effects as


59

a function of depth and decreasing vertical and horizontal resolution. The effective reservoir

parameters are obtained by averaging over defined depth intervals. The petrophysical logs

need to be upscaled, in order to model the expected resultant AVO seismic response.

AVO modeling has wide spread applications such as in seismic data acquisition design, and

pre–stack processing and interpretation which will go a long way to reduce the risks involved

in hydrocarbon exploration models and enhance reservoir characterization. AVO modeling

can be carried out over a wide range of approaches and data sets such as; Single interface

modeling, single – gather modeling, 2D and 3D stratigraphic modeling, 2D and 3D elastic

wave equation modeling (Yongyi et al., 2007).

This chapter will first present the theoretical framework and the assumptions used in

generating the different models. The target zones for modeling, data, and methodology are

also briefly described. The results obtained based on the assumptions used in generating the

different models are subsequently explained in the discussion.

4.2 Theoretical Background

4.2.1 Vp-Vs Relationships

The main input logs required for AVO modeling include Vp, Vs, and bulk density. Not all of

these petrophysical logs needed for modeling are usually available. Sometimes if present, they

may be in poor condition due to bore hole washouts which significantly affects log

measurements. The direct measurements of shear wave logs in particular is not common in the

older wells, and even for new oil field discoveries, it is not usually present in whole well

sections. As a result of these practical limitations, empirically derived Vs logs from measured

Vp logs such as the ‘’Mud Rock Equation’’ (Castagna et al., 1985) is used. This equation is

given by;

(4.1)

Where Vp and Vs are the compressional wave and shear wave velocities given in Km/s. This

equation has been derived for brine-saturated mudstones and shales, which tend to show a

linear relationship. Krief et al., (1990) also suggested a linear relationship (equation 4.2)

between Vp and Vs, with lithology dependent regression coefficients x and y.


60

(4.2)

Where Vp and Vs are the compressional wave and shear wave velocities given in Km/s. A

significant amount of work has been done by several authors, (Han et al., 1986, Dvorkin,

2008, Xu and White, 1996) to estimate the shear wave velocity from the Vp. The Vs log is

commonly used as an important frame indicator when accurately combined with Vp logs.

There is an inherent lithology dependence on the Vs estimates, in siliciclastic rocks.

Variations in the estimates can be observed due to variations in shape, grain size and

mineralogy (Mondol et al., 2009b, Mondol et al., 2010). However in this study the Castagna

et al. (1985) equation has been used to calculate Vs for shales due to its simple, yet robust

nature.

4.2.2 Gassmann fluid substitution

Fluid substitution has widespread applications in time lapse seismic reservoir monitoring and

also in AVO modeling and analysis. This technique has the potential to provide information

about insitu fluid scenarios and then model ‘’what if ‘’ scenarios. The practical application of

this technique is based on (Gassmann, 1951) given by equation 3;

(4.3)

Kframe, Kgrain, Kfluid represent, the effective drained framework modulus, bulk modulus of the

grains and fluids respectively. K and Φ represent the effective saturated bulk modulus and

porosity respectively. Gassmann fluid substitution offers a possibility to model the effective

seismic Vp, Vs and density of the target subsurface rock units, under different sets of

conditions such as; mineralogy, porosity, pressure, temperature, water salinity and pore fluid

saturation. However, in nature fluid substitution usually occurs gradually over time, yielding a

homogenous rather than a patchy saturation (Gelius and Johansen, 2010). The Gassmann

equation links the saturated rock bulk modulus to the porosity, fluid, and frame properties.

The main assumptions in Gassmann’s equation include;

An open pore system, in which all the pores are, connected (effective porosity). As a

consequence of this assumption, the fluid offers no resistance to shear deformation and


61

therefore the effective dry shear modulus is the same as the effective wet shear

modulus.

All the grains are assumed to have the same physical properties. By implication, the

properties of rocks with polymineralic phases are modeled into an effective

monomineralic phase. This can be achieved by using effective medium models, such

as the simple Voigt isostrain model. The Voigt average provides the upper bound to

the effective elastic properties of the different mineral phases.

Fully homogenous saturating pore fluid. This implies that the elastic properties of the

multiphase fluid system, is replaced by an effective fluid. This can be done using a

simple model such as Reuss isotress model, although more complex effective models

are available. A homogenous saturation implies that the fluid properties in the

individual pores are representative of the effective fluid properties.

Gassmann equation is valid for low frequencies. This is due to dispersion effects

associated with high frequencies. Frequency dependent fluid oscillations cause

velocities to vary as a function of frequency.

Gassmann’s theory has a limitation with regards to shaly sequences due to the presence of

electrostatically bound water, in unconnected pore systems.

4.2.3 Synthetic Seismogram

From a seismic acquisition point of view, a seismic trace is a time measurement

corresponding to a given source-receiver pair. The distance between any given source receiver

pair, is referred to as offset. In order to model the effective rock parameters from given

petrophysical logs, a synthetic seismogram can then be generated. This synthetic seismogram

can then be compared to real seismic data. The main input required to generate a synthetic

seismogram are the density, Vp logs and an assigned wavelet. The acoustic impedance (z) of a

medium is given by a product of the density and Velocity (Vp) of that medium.

Fig. 4.1 Simple two layer model with contrasts in acoustic impedance (Z).


62

(4.4)

The impedance contrast across an interface is responsible for the wave reflection phenomena.

The reflection coefficient (R), across an interface is given by

(4.5)

From the formular above, the magnitude of the reflection coefficient depends on the constrast

in acoustic impedance. Mathematicaly a seimic trace can then be modeled as a linear

convolution between a wavelet and the earth reflectivity series as shown in Figure 4.2. The

earth’s reflectivity series can be described as a time series of spikes, each of which actually

represents a zero offset plane wave reflection coefficient.

(4.6)

Fig. 4.2 Convolution between the wavelet and the reflectivity series (adapted from Mondol,

2010).

Some assumptions in the convolutional trace model include;

No variation of the source pulse s(t), with depth i.e. stationary source pulse

No noise contribution present

Normal incident plane waves through a simple horizontal stratigraphically layered

earth model.

The synthetic seismogram employed in this study, compirise of NMO corrected Common Mid

Point (CMP) gathers (groups of traces with a common mid-point between source and

recievers). Analysis of these gathers usually show offset dependent reflectivity (variation in

the zero-offset reflection coefficient as a function of angle of incidence).


63

4.2.3 Angle dependent reflection coefficient

A normal incident P-wave on any given interface usually generates only reflected and

transmitted P-waves. In cases where the angle of incidence Φ1>0, a second pair of reflected

and transmited S-waves are generated. This phenomenon is known as mode-conversion as

shown in Figure 4.3.

Fig. 4.3 Mode conversion of P-waves (modified from Mondol, 2010).

The waves across the interface between both media are reflected and refracted according to

Snell’s law given by ;

(4.7)

Conventional marine seismic acquisition surveys with streamers located at the sea surface do

not record these converted S-waves. Despite this limmitation, the P-P reflection coefficient

contains indirect shear wave information as a result of mode conversion. AVO is the

appropriate technique that tries to extract the hidden information.

Knot-Zoeppritz equations describe the variation in reflection coefficient as a function of

angle of incidence. The exact Zoeppritz equations do not provide a simple physical

interpretation which can be applied practically. There are several approximations to the

Zoeppritz equations. Aki and Richards (1980), proposed a first order linear approximation of

the Zoeppritz equation. For small angles (best fit to Zoeppritz up to approximately 35ᶿ), the

linearised version is simplified according to Wiggens approximation into ;


64

(4.8)

Where;

(4.9)

(4.10)

(4.11)

Rpp(Φ) is the P-P reflection coeffecient at angle of incidence(Φ). Rp and B in AVO

terminology are refered to as the AVO intercept and AVO gradient respectively. Rp and Rs

are the zero offset reflection coefficients for Vp and Vs respectively, after performing a

linearized first order analysis. ΔVp and ΔVs represents the velocity contrast across the

interface, meanwhile Vp and Vs represent the average velocity across the interface. Δρ

represents the density contrast across the interface, meanwhile ρ is the average denisty across

the interface.

The Aki-Richards approximation is simple, yet robust enough to be applied practically in

AVO synthetic modeling work flows. The AVO quantities can be plotted seperately as a

gradient stack or an intercept stack. Depending on the purpose, combined sections can also be

constructed which can enhance bright spot events at the expense of lithological events. On the

other hand, the combined sections could be used in a manner so as enhance strong shear wave

reflectivities (Gelius and Johansen, 2010).

4.2.4 Classification of reservoir sands based on AVO

The seismically extracted AVO parameters (AVO intercept and AVO gradient) have been

used by to classify gas sands (Rutherford and Williams, 1989). This classification was

initially based just on the AVO intercept, taken as zero-offset reflection coeffcient. Modern

classification schemes combine both the AVO gradient and AVO intercept, giving rise to

four gas sand classes in a simplified overlying shale and sand reservoir ;

Class I gas sands: These are sands with impedances higher than the overlying shale.

They show a large positive zero offset reflection coefficient at the boundery between


65

shale and sand. The reflection coefficients decreases with offset. This type of

behaviour is common for highly compacted mature sands in onshore environments.

Class II gas sands: These are sands with a small impedance contrast with the

overlying shale. As a result they have very low normal incidence reflectivity. These

sands usually show intermediate levels of mechanical and chemical compaction. A

large change in reflecitivity with offset is typical of this class, and in some cases

polarity changes could occur if the zero offset reflection coefficient is positive. This

class is common to both offshore and onshore sands.

Class III and IV gas sands: These gas sands have a lower impedance than the

overlying shales. This is common for unconsolidated sands, and show a large negative

zero-offset reflectivity. These sands are classical ‘’bright spots’’ on stacked seismic

data, because they have large reflectivities for all ofsets. Both class III and IV sands

are usually associated with a marine envirionment. Class IV sands show decreasing

reflectivity with offset (positive gradient), meanwhile class III shows an inverse

relationship as shown in Figure 4.4.

Fig. 4.4 Rutherford and Williams (1989) classification scheme based on the AVO intercept.

From the above classification it is evident that the reflectivities of gas sands do not necessarily

increase with offset. More advanced classification schemes combine the AVO gradient and

the AVO intercept as shown in Figure 4.5 .


66

Fig. 4.5 AVO intercept and AVO gradient crossplot classification of gas sands (Rutherford

and Williams,1989).

The background trend represents the gradient-intercept trend for a water saturated siliciclastic

system, based on the ‘’mud rock’’ line by (Castagna et al., 1985). The background trend line

varies depending on the Vp/Vs ratio. By combining Rutherford and Williams (1989)

classification scheme, together with the ‘’mud rock’’ line, anomalous data points falling far

from the background values can easily be iditified as potential candidates for gas sands. The

AVO intercept depends on the impedance constrast, meanwhile the AVO gradient depends on

contrasts in the poisson’s ratio. A summary of the behaviour of the various gas sands has been

presented in Table 4.1.

Table 4.1 Summary Rutherford and Williams classification scheme assuming a

‘’background’’ trend with a negative slope (Castagna et al., 1998).

CLASS RELATIVE

IMPEDANCE

QUADRANT A B REMARKS

I Higher than

overlying unit

IV + - Reflection coefficient( and

magnitude)decrease with

increasing offset

II About the same as

the overlying unit

III or IV ± - Reflection magnitude may

increase or decrease with

offset, and may referse


67

polarity

III Lower than

overlying unit

III - - Reflection magnitude

increases with offset

IV Lower than

overlying unit

II - + Reflection magnitude

decreases with offset

4.3 Database and methodology

AVO modeling for ‘in-situ’ and ‘what if’ scenarios has been carried out for two main

reservoir units; the thicker, and deeper Kobbe oil reservoir of Mid – Triassic age, and the

Tubåen reservoir of Jurassic age which is thinner, shallower and with a small gas cap. Both

reservoirs have different cap rock properties. The Fuglen cap rock shows a relatively higher

resistivity, Vp, Vs, and bulk density compared to the base of the Snadd Formation which acts

as the cap rock for the Kobbe reservoir (Figure 4.6). In order to understand the AVO response

of these reservoir units, it is important to investigate the lateral consistency of the AVO

response for each of these reservoirs throughout in the Goliat Field based on the available

well data. In order to achieve this, a suite of 6 wells have been used.

The main input wells for AVO modeling for the Kobbe and Tubåen reservoirs used in this

study are presented in Table 4.2 and Table 4.3. The Kobbe reservoir is deeper than the total

drilled depth in wells 7122/7-1 and 7122/7-2. The Tubåen reservoir pinches out laterally

between the wells into the Fruholmen formation in well 7122/7-4 and well 7122/7-5. Well

7122/7-5A is a side track well from well 7122/7-5, and does not have measured log values for

the Tubåen reservoir. Due to a combination of lateral facies variation and incomplete log

suits, the same reservoir cannot be analyzed throughout the available well data.

Table 4.2 Kobbe formation depth and thickness variation.

WELL NAME DEPTH(M) BSF THICKNESS (M)

7122/7-3 1440 236

7122/7-4 1399 248

7122/7-5 1474 258


68

Table 4.3 Tubåen formation depth and thickness variation.

WELL NAME DEPTH(M) BSF THICKNESS (M)

7122/7-1 696 24

7122/7-2 683 118

7122/7-3 719 93

The Kobbe reservoir (Table 4.2) is located at much deeper depths compared to the Tubåen

reservoir (Table 4.3), though both are not presently located at maximum burial depths, due to

regional exhumation in this area. Prior to exhumation these reservoirs had been subjected to

sufficient burial depths and corresponding temperatures for cementation to take place, which

significantly reduces the reservoir quality and increases the degree of heterogeneity. The base

of the Snadd formation is characterized by thick shale units in this area which serves as cap

rock for the underlying Kobbe reservoirs as shown in Figure 4.6.

From Figure 4.6, it is evident that there is a larger contrast in resistivity between the base of

the Snadd Formation, and the underlying Kobbe reservoir, compared to the case of Fuglen cap

rock and the Tubåen reservoir. The computed poison’s ratio for the highlighted target

horizons shows a significant drop. The target for AVO modeling carried out in this study

focuses on the top section of the reservoir, at the interface between the cap and reservoir rock.

These target interfaces show a relatively clear contrast based on the gamma ray log.

Relatively clean sands were chosen to minimize limitations in the Gassmann model with

regard to shaly sandstones.


69

Fig. 4.6 Gamma, Deep Resistivity, Density, Vs, Vp and Computed Poisson log intervals for

the target zones (highlighted in yellow) in the Tubåen and Kobbe reservoir (well 7122/7-3).

A practical approach to AVO modeling using Hampson-Russel software involves generating a

model for in-situ and then ‘’what if ’’ scenarios. AVO modeling is usually carried out in order

to determine the anticipated anomaly.

The main input logs used in AVO modeling are Vp, Vs and density logs. Direct recordings of

Vs are usually not present in all the wells, and the Vp log was then used to create a Vs log

using linear Log transforms in Hampson-Russel. The Castanga ‘’mud rock’’ equation was

then applied to generate Vs in the wells where this log is absent. An appropriate well log

upscale was then chosen depending on the size of the target reservoir. Upscaling is important

because the well logs contain higher frequency information than actual seismic data. This also

reduces the computing time in cases of a large data set. There is usually a tradeoff between

the resolution and increasing the block size. Larger block sizes will contain less frequency

information. The maximum block size used in this study is 25m. The corresponding computed

impedance and computed reflectivity are automatically generated depending on which of the

input upscaled logs are active.


70

One important step in the work flow needed to generate a synthetic seismogram, involves

creating a seismic wavelet. The default Ricker Linear wavelet was used in this study.

However for comparison to real seismic data, much more complex forms of wavelet

extraction is necessary either from the suite of well logs or from the seismic alone.

Fig. 4.7 Time and frequency domain of the Ricker linear wavelet used in this study.

A peak using this wavelet signifies an increase in acoustic impedance. The domain frequency

of this wavelet as shown in Figure 4.7 is 45 Hz. The Ricker linear wavelet used has a

wavelength of 200ms and employs a sample rate of 2ms. The average phase of this wavelet is

a zero phase. There are no side lobes in this wavelet, leading to an ideal signal-to-noise ratio.

This gives an exaggerated vertical resolution than can be normally achieved in a seismic

exploration survey.

In order to generate an offset/angle dependent synthetic seismogram, automated ray tracing

was used to calculate the angle of incidence. The amplitudes were then calculated using the

full Zoeppritz equation and then analyzed using the simplified Aki-Richard equation. The

offset range used in generating the synthetic seismogram ranged from 0-1000m, with an

output sample rate of 2ms. No effects of geometrical spreading or transmission losses were

considered in the model. The synthetic seismic outputs used in this study are NMO corrected

CMP gathers. Output reflectivity was chosen over output amplitude in Hampson-Russel, as

this generates a synthetic seismic with better vertical resolution (Figure 4.8).


71

Fig. 4.8 Resolution differences using output ‘’reflectivity’’ compared with output

‘’amplitude’’ for Kobbe reservoir in well 7122/7-3.

The Biot-Gassmann method is then applied using Fluid Replacement Modeling (FRM) in

Hampson-Russel to investigate ‘what if’ scenarios for different fluid types and saturations.

The frame properties of the rock are assumed to be constant. Pressure effects are not modeled

in this study. As a consequence of this, the input porosity is set to be equal to output fluid

substituted model.

Table 4.4 Matrix properties used for fluid replacement modeling.

Matrix Type Sandstone

Bulk Modulus 40 (GPa)

Shear Modulus 44 (GPa)

Density 2.65 (g/cc)

Table 4.5 Fluid properties used for fluid replacement modeling.

Fluid Type Bulk Modulus(GPa) Density (g/cc)

Brine 2.38 1.09

Oil 1 0.75

Gas 0.02 0.1

The matrix and fluid properties presented in Table 4.4 and 4.5 represent the default values in

the Hampson-Russel software based on Batzle and Wang (1992). These properties were set as

constant and used for the different fluid saturations in this study. The assumed matrix and


72

fluid properties, used for different reservoir sections is an over simplification of the actual

properties. Despite these simplifications, the Biot-Gassmann approach is yet robust enough to

give consistent results.

4.4 Results

The two main reservoirs of focus are the Tubåen and Kobbe formations. The input logs shown

in Figure 4.9 have been averaged for every 15m (block 15). The density, Vs and Vp upscaled

logs of the target zone are highlighted in yellow. The offset dependent reflectivity for all the

wells has been generated using the Zoeppritz equation. The amplitudes have been extracted

using the Aki-Richard two term parameter equation, and only the best fit lines have been

represented.

Fig. 4.9 Density, Vs and Vp logs generated by 15 m averaging for Tubåen (A) and Kobbe (B)

reservoirs.

4.5 Sensitivity analysis

Despite the simplicity of the Gassmann model, it is still robust enough in providing

quantitative changes in petrophysical rock parameters for different fluid scenarios for the

Kobbe reservoir in well 7122/7-3. The insitu modeled fluid saturation is an oil filled reservoir.


73

Table 4.6 Variation in Vp, Vs, density and Poisson’s ratio with changing gas saturations.

Gas (%) %ΔVp %ΔPoissons Ratio %ΔDensity %ΔVs

10 -17.247 -29.637 -0.117 0.059

50 -21.573 -43.369 -0.588 0.296

90. -22.043 -45.723 -1.059 0.534

The analysis in Table 4.6 compares the sensitivity of the density, Poisson’s ratio, Vs and Vp

in a gas–oil system. After introducing 10% Gas into the reservoir, there is a significant drop in

the Poisson’s ratio and Vp by 29.64% and 17.25% respectively as shown in Figure. 4.10. For

the same 10% gas saturation, there is a relatively insignificant drop in the density, meanwhile

Vs shows a rather slight increase as expected.

Fig. 4.10 Effect of changing gas saturations on Vp, density, Vs and Poisson’s ratio.

Increasing the gas saturations after 10%, results to much smaller changes in these

petrophysical rock parameters when compared to incipient introduction of gas. This pattern is

the same for increasing gas saturations up to 90% gas. Vs is relatively insensitive to the

changing saturations and can be used as frame indicator.


74

Fig. 4.11 Synthetic seismic generated using a Ricker linear wavelet.

The synthetic seismic for the different saturation scenarios, have been generated using a

Ricker Linear wavelet and applying the Zoeppritz equation for a total offset of 1000m (Figure

4.11). The data has been displayed using the normal polarity convention. The wiggled traces

have a color infill for the positive reflection coefficients. The corresponding synthetic seismic

for the different saturations show a slight change just for the initial 10% gas saturation. The

synthetic seismic data for higher gas saturations are more or less the same. This observation is

consistent with the relatively large changes in the petrophysical parameters just for incipient

10% gas saturation.


75

Fig. 4.12 Effect of changing gas saturations on the zero-offset reflectivity (Rp) in an oil-gas

system.

The effect of changing saturations is much more evident, by comparing the angle dependent

reflectivity with offset (Figure 4.12). The magnitude of the zero offset reflection coefficient

(AVO intercept), increases i.e. becomes more negative with increasing gas saturation. All four

scenarios have a negative AVO intercept and a negative gradient. The percentage change in

the AVO intercept (calculated for each successive increment in gas saturation), is greatest for

10% gas and much smaller for 50% and 90% respectively.

4.5.1 Variations in half space models

The models described here represent simple two layer models with a single interface. Only the

top target reflector has been used in this study without considering the base of the reservoir

horizon of interest.

The magnitude in the zero offset reflection coefficient, ranges from - 0.050 to -0.009 for all

fluid scenarios in the Tubåen and Kobbe Formations considered in this study (Table 4.7 and

Table 4.8). Subtracting the magnitudes of both limits and dividing by 2 gives a ‘’cut-off

‘value of -0.021. Rp values more negative than this value are considered in this study as

relatively strong, meanwhile amplitude values less negative than the cut-off limit are

considered weak. The cap rocks for the Tubåen and Kobbe reservoir units are Fuglen and

Snadd Formations.


76

Fig. 4.13 Angle dependent reflectivity for Tubåen Reservoir in an oil-gas system.

Table 4.7 Tubåen Reservoir AVO classification.

WELL PORE FLUID CLASS ZERO OFFSET Rp

7122/7-1 INSITU OIL WEAK IV -0.009

GAS STRONG III -0.032



7122/7-3 INSITU GAS STRONG IV -0.030

BRINE STRONG IV -0.031

The Tubåen reservoir shown in Figure 4.13 shows a negative zero offset reflectivity with

offset and a positive gradient (class IV) for all three wells. However, there are variations

observed in the zero offset reflectivity. Insitu gas scenario in well 7122/7-3 shows the largest

zero offset reflectivity compared to the other insitu oil scenarios. After Gassmann fluid

replacement modeling (FRM), the insitu oil scenarios, change from weak to relatively strong

zero offset reflectivity, and with a negative gradient (becomes more negative with increasing

offset). The increase in reflectivity with offset is greatest for the gas model in well 7122/7-1.

Meanwhile the change in reflectivity with offset for insitu oil scenario for the Tubåen


77

reservoir in well 7122/7-2 is greater than that observed for the corresponding gas model in the

same well. There is no significant change after replacement of insitu gas with brine in well

7122/7-3 as shown in Figure 4.14.

Fig. 4.14 Quantitative changes in Rp for the Tubåen reservoir after fluid replacement

modeling.

All insitu models for the Kobbe reservoir based on the ‘’cut off value’’ show weak zero offset

reflectivity. The insitu oil models for the Kobbe reservoir in well 7122/7-3 and 7122/7-4 show

a positive gradient and negative AVO intercept (Class IV), same with the insitu brine scenario

in well 7122/7-5A. Only well 7122/7-3 insitu oil scenario shows a negative gradient. For the

insitu scenarios, there is a progressive increase in the strength of the zero offset reflectivity

from brine to oil and gas. The zero offset values for insitu Kobbe oil reservoir lies between

those of brine and gas. The insitu oil scenario in well 7122/7-3 shows the largest change in

reflectivity with offset.

After fluid replacement modeling, the Kobbe reservoir in all the wells, show a stronger zero

offset reflectivity, with negative gradients as shown in Figure 4.15. The gas model

corresponding to in situ brine shows the strongest zero offset reflectivity, when compared to

all modeled gas scenarios in the different wells. Well 7122/7-3 Kobbe gas scenario shows

more negative zero offset reflectivity than its insitu counterpart but maintains the same

negative gradient. All the other wells show a change in the sign of the gradient, after fluid

substitution.


78

Fig. 4.15 Kobbe reservoir offset dependent reflectivity before and after fluid substitution.

Table 4.8 Kobbe Reservoir AVO classification.

WELL PORE FLUID CLASS ZERO OFFSET Rp

7122/7-3 INSITU OIL WEAK III -0.018




7122/7-5 INSITU BRINE WEAK CLASS IV -0.009

GAS STRONG CLASS III -0.050

7122/7-5A INSITU OIL WEAK CLASS IV -0.018

GAS STRONG CLASS III -0.040

The change in reflectivity (Rp) between the insitu scenarios and the corresponding gas model

for the Kobbe reservoir is seen to vary depending on insitu fluid type illustrated in Figure

4.16. The greatest change is observed for insitu brine scenario in well 7122/7-5. The

corresponding changes in reflectivity for the insitu oil models are relatively same.


79

Fig. 4.16 Quantitative changes in Rp for the Kobbe reservoir after fluid replacement

modeling.

4.5.2 Effect of block size variation on the AVO signature

The synthetic NMO corrected CDP gathers generated using a Ricker, linear wavelet for the

Tubåen reservoir in well 7122/7-3 shows distinct results for different block sizes (Figure.

4.17). Block 25 and Block 15 synthetics are based on averaged Vp, Vs, density logs for every

25m and 15 m respectively.

Fig. 4.17 Variation in synthetic NMO corrected CMP gathers with block size 15 and 25.

Block 25 shows data with a lower resolution and more noise introduced into the data. The

reflection corresponding to top Fuglen and Tubåen occur at different depths depending on the


80

averaging size used. This has consequences in the resulting AVO response for the target

reflectors.

Fig. 4.18 Effect of block size on the AVO gradient.

Block 25 shows a negative gradient for both gas and brine models, meanwhile Block 15

shows a positive gradient for both brine and gas model scenarios (Figure 4.18). There is no

significant change in the magnitude of the reflectivity for both block sizes before and after

fluid substitution.

4.5.3 Kobbe and Tubåen angle dependent reflectivity comparison

The Tubåen reservoir is located at shallower depths (719m) than the Kobbe reservoir at

1440m BSF. The Fuglen cap rock to the Tubåen reservoir has different properties compared

to the shales of the base Snadd Formation which serve as the cap rock for the Kobbe reservoir

(Figure 4.6). The insitu fluid scenario for the Tubåen reservoir comprises of a gas-oil system,

meanwhile the Kobbe reservoir is insitu oil filled. Both reservoirs fall within the zone affected

by quartz cementation, as defined from the Vp-depth trend in this same well (7122/7-3) in

chapter 3. Both reservoirs after 100% gas substitution show negative zero offset reflectivity.

The Tubåen reservoir shows a typical class IV positive gradient, meanwhile Kobbe reservoir

shows reflectivity increasing with offset characteristic of a class III as seen in Figure 4.19.


81

Figure 4.19 Gas scenario reflectivity versus offset for top Tubåen and Kobbe reservoir.

4.6 Discussion

4.6.1 Sensitivity study

The effect of fluid substitution on Vp, Vs, density, and Poisson’s ratio in the Kobbe oil

reservoir in well 7122/7-3 shows a consistent and interdependent relationship between these

petrophysical properties. The Vp/Vs ratio determines the magnitude of the Poisson’s ratio.

(4.12)

(4.13)

K, µ, and ρ represent the effective saturated bulk modulus, shear modulus and density

respectively. From equations 4.12 and 4.13, Vp and Vs will both increase with a decrease in

effective density, if all other parameters are kept constant. The shear modulus is insensitive to

fluid type and saturations, such that the shear modulus of rock frame is same as the shear

modulus of the same rock saturated with fluid. Therefore the only varying parameter


82

controlling Vs is the density. The fluid properties for gas and oil used in this oil-gas two phase

system are shown in Table 4.5. The density of gas is lower than density of oil, but the bulk

modulus is the controlling factor determining Vp. Gas has much more lower compressibility

compared to oil. As a result just a 10% gas input into the oil system, leads to a large

percentage drop in Vp of 17.25%. This large change is due to a correspondingly large change

in the saturated bulk modulus (Ksat) of the medium.

Table 4.9 Density and velocity cross plot with increasing gas saturation.

%Gas Density(kg/m3) Ksat (Mpa)

0 2561.87 14800

10 2558.86 12200

50 2546.8 11800

90 2534.74 11600

A cross plot of density and the effective saturated bulk modulus (Ksat) in Figure 4.20 shows a

consistent decrease in both parameters with increasing gas saturation. The effective fluid bulk

modulus used as input in the Gassmann model for the different saturations, is calculated using

the Reuss harmonic averaging in Hampson-Russel. This follows the assumption of a

homogeneous saturation. However, a much larger decrease is observed for Ksat at the 10%

gas saturation. Increasing the gas saturation after 10%, shows very low sensitivities in Ksat

compared to density. The percentage change in Vp is higher for 50% and 90% Gas saturation,

because these saturations were all compared to the insitu scenario as a reference.

Fig. 4.20 Cross plot between the saturated bulk modulus and density color coded with gas

saturation.


83

However it is evident that adding the gas saturation in the oil-gas system from 50% to 90%,

leads to a difference of less than 2% change in the corresponding velocity at 90% gas

saturation.

The effective bulk density consistently drops with increasing gas saturation. The model uses

Voigt arithmetic average to calculate the effective bulk density. Meanwhile the Reuss

harmonic average is used in calculating the effective fluid bulk modulus used in the

Gassmann equation. Arithmetic averaging is used for the bulk density because density is a

scalar quantity with no directional dependence.

Vs however shows a slight increase with increasing gas saturation. The percentage increase in

Vs is roughly proportional to the percentage decrease in density. The Poisson’s ratio shows

the largest drop with increasing gas saturation. This is just due to the large change in Vp and

an almost insignificant change in Vs on addition of gas into the system. When the percentage

change of all four petrophysical parameters are compared using the same scale, their

sensitivity can be ranked in the order; Poisson’s ratio> Vp > density > Vs.

The corresponding synthetic seismic data for all three gas scenarios and the reference insitu

oil scenario are shown in Figure 4.11. There is a slight increase in the strength of the internal

reservoir reflectors. After introducing gas into the reservoir, the acoustic impedance contrast

increases which also causes an increase in the strength of the reflectors. This change is

observed just for the initial 10% gas saturation. With increasing gas saturations, the

corresponding synthetic seismics look the same. This is largely due to the decreasing

sensitivity of the effective saturated bulk modulus and Vp for progressively higher gas

saturations. The percentage change in Vp is much larger than Vs and density and tend to

control the observed synthetic seismic and AVO response for the different saturations.

However negative reflection coefficients, show no apparent change in the reflection strength,

just because the data is displayed in the normal polarity convention, in which the negative

wiggled trace is not filled.

The top reservoir target reflector shows clearer differences with changing fluid saturations,

when comparing the zero offset reflectivity for the different gas saturations. The change in

reflectivity as expected is largest for the initial 10% gas. With increasing gas saturation, the

reflection coefficient increases in magnitude because, the acoustic impedance contrast

between the reservoir and the cap rock also increases.


84

4.6.2 Variation in half space models

The results shown in Figures 4.13 and 4.15 indicate that fluid substitution in both the Tubåen

and Kobbe reservoirs has a significant influence on both the AVO intercept and the AVO

gradient. The magnitude of the AVO intercept is controlled by the contrast in acoustic

impedance meanwhile the AVO gradient is controlled much more by changes in the Poisson’s

ratio.

4.6.2.1 Tubåen AVO response

For the Tubåen reservoir, substituting oil for gas in well 7122/7-1 and 7122/7-2 causes an

increase in the strength of the zero offset reflection coefficient (becomes more negative)

(Figure 4.13). The presence of less dense gas reduces the impedance of the reservoir and

increases the acoustic impedance contrast between the Fuglen cap rock and the underlying

reservoir. Due to the dependence of Vp on the bulk modulus, the Poisson’s ratio also

decreases significantly in the presence of gas in both reservoirs (Figure 4.21).

Fig. 4.21 Change in Poisson’s ratio for Tubåen reservoir after fluid substitution.

The AVO gradient based on a first order linearized simplification of the Zoeppritz equation is

given by:

(4.14)


85

From equation 4.14, it is evident that a decrease in the Poisson’s ratio (Vp/Vs) leads to an

increase in (Vs/Vp) 2

and causes the AVO gradient to become more negative. The increase in

reflection coefficient with angle/offset (gradient) is stronger for the gas model in well 7122/7-

1 than the corresponding gas model in well 7122/7-2. This can be correlated to the change in

Poisson’s ratio as a result of fluid substitution. The change in the Poisson’s ratio for both the

insitu oil scenario and the corresponding gas model is shown in Figure 4.21. The change in

Poisson’s ratio is larger for well 7122/7-1 compared to 7122/7-2 and explains why the

gradient is stronger. Based on the gradient, well 7122/7-1 gas model is a class III (negative

gradient) Tubåen reservoir; meanwhile in well 7122/7-2 Tubåen reservoir shows a positive

gradient typical for class IV sands. Castagna et al. (1998), showed the contribution of Vp, Vs

and density on the gradient of class III and class IV sands. Based on this study, contribution

from the density contrast leads to positive gradients. Meanwhile a contribution from ΔVp

leads to a negative gradient. However it was concluded that the key parameter controlling the

gradient with offset is ΔVs. Combining this information together with Aki and Richards

(1980) approximation, it is then possible to explain the difference in the positive and negative

gradient for the Tubåen gas models.

(4.15)

Where P is the ray parameter given by (Sin θ/Vp) and θ is the average of the angle of

incidence and refraction. The remaining terms are linearized described in equations 4.10 and

4.11.

Table 4.10 Change in Vs for Tubåen reservoir.

Well and Pore Fluid Reservoir (Tubåen)

Vs (m/s)

Cap Rock (Fuglen)

Vs (m/s)

(ΔVs)

7122/7-1 Insitu oil 1710.17 1800.36 -

Gas model 2024.23 +

7122/7-2 Insitu oil 1540.43 1676.22 -

Gas model 1810.84 +

7122/7-3 Insitu gas 1722.55 1289.49 +

Mode conversion of Vp into Vs occurs only for angles of incidence greater than zero. This

implies that ΔVs has no effect on the zero offset reflection coefficient. From equation 4.15, a


86

positive ΔVs (increase in shear velocity across the cap rock-reservoir interface) increases the

shear contribution for increasing offsets, leading to an increase in the total amplitude with

offset. This leads to a more class III type signature. A reverse scenario occurs for a negative

ΔVs.

For the Tubåen reservoir, the insitu oil models in well 7122/7-1 and 7122/7-2 show a decrease

in shear velocity across the cap rock- reservoir interface (Table 4.10). This correlates well

with the corresponding class IV signatures for this scenario. After substituting oil for gas,

ΔVs becomes positive and correlates with an increase in reflection coefficients at higher

angles/offsets in a class III type response. However, a classical class IV response is usually

anticipated when a high velocity unit (‘’tight cap rock’’) such as the Fuglen cap rock overlies

a porous reservoir. This is the case observed insitu in well 7122/7-3, using direct shear wave

measurements as opposed to the other wells using the linear transform from the measured Vp

according to Castagna et al. (1985).

4.6.2.2 Kobbe AVO response

There is a change of class observed for all the Kobbe insitu scenarios in Figure 4.15 except

for well 7122/7-3, after fluid substitution. The insitu brine case in well 7122/7-5 shows the

smallest zero offset reflection coefficient. This is due to a lower acoustic impedance contrast

with the overlying cap rock which is brine filled. The corresponding gas model for this well

shows the strongest zero offset reflection coefficient. This is consistent with the fact that, the

acoustic impedance contrast for brine and gas larger than that for oil and gas as is the case in

the other wells. The base of the Snadd formation, which serves as the cap rock for the Kobbe

reservoir, is not a ‘tight cap’ rock compared to the Fuglen cap rock. The introduction of gas,

lowers the Poisson’s ratio in all the wells, and causes both the AVO intercept and AVO

gradient to become more negative. Except for well 7122/7-3, all the other wells show ‘’less

positive gradients’’ after fluid replacement with gas. The positive ΔVs observed for all the gas

models after fluid substitution (Table 4.11), is also consistent with equation 4.15, leading to

an enhanced total amplitude increase with offset. Well 7122/7-4 shows the strongest gradient

compared to the insitu oil model. This is consistent with the largest observed Poisson’s ratio

change for well 7122/7-4. On the other hand well 7122/7-3 in Figure 4.22 shows the smallest

change in Poisson’s ratio, and therefore very similar gradients before and after fluid

substitution.


87

Fig. 4.22 Change in Poisson’s ratio for Kobbe reservoir after fluid substitution.

Table 4.11 Change in Vs for Kobbe reservoir.

Well and Pore Fluid

Reservoir (Kobbe)

Vs (m/s)

Cap Rock (Snadd)

Vs (m/s)

(ΔVs)

7122/7-3 Insitu oil 1724.19 1497.96 +

Gas model 1765.37 +

7122/7-4 Insitu oil 1557.16 1528.63 +

Gas model 1836.85 +

7122/7-5 Insitu brine 1596.58 1499.48 +

Gas model 1671.19 +

7122/7-5A Insitu oil 1661.68 1614.74 +

Gas model 1907.06 +

4.6.2.3 Tubåen and Kobbe AVO gas response and the effects of block size

variation

Differences in the strength of the cap rocks are known to influence the AVO response of the

underlying reservoirs (Lars et al., 2006, Avseth et al., 2008). The difference in class III and IV

for Kobbe and Tubåen respectively in Figure 4.19 may be due to a combined effect of

compaction and differences in the cap rock. The Fuglen formation is a ‘’tight cap rock’’

unlike the Snadd formation. Similar observations for ‘tight cap rocks’ have been observed by

Castagna et al. (1998). Despite this, the zero offset coefficient for the Snadd gas model, is


88

larger than for the Tubåen reservoir. This is possibly due to the fact that the model uses frame

properties of pure quartz.

The target reflector (B) for the insitu Tubåen reservoir in Figure 4.23 is seen as a strong

negative reflector and can be followed in block 15, but this same reflector is averaged out in

block 25. This explains the difference in the AVO response. The zero offset coefficients for

the top Tubåen is almost zero for a gas model.

Fig. 4.23 NMO corrected synthetic CDP gathers for block 25 and block 15 for top Tubåen

reservoir in well 7122/7-3.

The modeled reservoir interval is too thin to be observed in the seismic due to the large

averaging used. Top Fuglen (cap rock) is seen as a strong positive reflector and can be

correlated across both synthetic sections. Amount of noise in the data is increasing with

increasing block size. This is due to a reduction in resolution with increasing block size. The

phase change observed in block 25 may be the result of effects of NMO stretching for large

offsets for the top Fuglen reservoir in block 25. This study puts more emphasis on the zero

offset reflection coefficient, and the AVO gradient for near offsets, to minimize possible

effects of NMO stretching in the synthetic seismic section.

4.7 Uncertainties in the modeled scenarios

Due to the absence of core data there was no control on the mineralogy. Differences in the gas

models obtained using Gassmann fluid substitution and the insitu gas model in well 7122/7-3


89

may arise due to over simplifications made in the model. The frame properties were set to be

equal to that of pure quartz. These assumptions impose a limitation in the interpretation of the

AVO response. This is because observed reservoir heterogeneity from the gamma ray log has

not been taken into account as shown in Figure 4.24.

Fig. 4.24 Gamma ray logs for Tubåen (A) and Kobbe (B) reservoirs.

These siliciclastic reservoirs have differential amounts of shale even for the same reservoir in

the different wells. The presence of shale increases the Vp/Vs ratio, which influences the

AVO gradient. In addition, shales usually have unconnected pores, which violates

Gassmann’s assumption of total communication in the pore space.

The pressure dependence on the output for fluid substitution has not been taken into account,

due to the absence of data to constrain the different pressure regimes in both the Kobbe and

Tubåen reservoirs, located at different depths. As a result, the input porosity is set equal to the

output porosity, and this distorts the effect of compaction on the AVO response. The effects

of biodegradation have also not been taken into account. The oil in the Kobbe reservoir is not

biodegraded (NPD Factpages), unlike the hydrocarbon accumulations of the Tubåen

formation. Biodegraded oils are usually denser than non biodegraded oils. This difference

may also pose limitations to the interpretations discussed in this study.

Chapter 5. Summary and conlusion

90

CHAPTER 5 SUMMARY AND CONCLUSION

5.1 Summary and conclusion

The Barents shelf area has recorded a higher success rate than the Norwegian North Sea in

terms of initial frontier exploration efforts. Multiple source rocks are also found at different

stratigraphic intervals, including the wide spread Upper Jurassic-Cretaceous Hekkingen

Formation which is equivalent to the Kimmeridge shale unit in the North Sea. The distance to

potential markets, climate, dominant gas products, large water depths, and the Cenozoic uplift

pose a lot of challenges for frontier exploration, development and production from the

discovered oil and gas fields.

The PDO approved Goliat oil field located on the Finnmark platform, is one among a series of

rather few oil finds among the major gas reserves such as the Snøhvit and Shtokmann gas

fields. The occurrence of non-cogenetic gas in this field is a good sign of a live petroleum

system in the area. Uplift had devastating consequences not just to the source rocks, but also

the reservoir and cap rocks in this area. The main reservoir intervals include; the Tubåen,

Fruholmen, Snadd, Kobbe, Klappmyss, Havert Formations. However this work lays emphasis

on the Kobbe and Tubåen reservoirs including their cap rocks (base Snadd and Fuglen

Formations respectively). The major structure cutting through this field is the Troms

Finnmark Fault Complex (TFFC). The main structural trap is a roll over anticline.

Compaction trends in the Goliat have been investigated by comparing a suite petrophysical

well log data, and experimental laboratory compaction trends. Different clay – clay and clay –

silt curves have been utilized. Generally as a function of depth, Vp and density increase

meanwhile the porosity reduces as expected. The transition from mechanical to chemical

compaction for siliciclastic rocks has been investigated using two techniques. First by

recognizing an abrupt increase in the Vp – depth trend within the same lithology (inferred

from the gamma ray log. Incorporating information from a shear modulus-porosity cross plot

also helped to constrain the depth at which this transition occurs. There are uncertainties in

the paleogeothermal gradient, and the actual mineralogy of the rock units discussed in this

work.


91

AVO fluid replacement modeling of the Kobbe and Tubåen reservoir provides additional

insight into the rock properties of these reservoir rocks. Despite the major limitations and

simplifications in the Gassmann fluid replacement model, it still provides a robust approach.

The AVO signature is not dependent on the properties of the reservoir alone but also on the

overburden through which the seismic pulse propagates. As a result, fluid replacement

modeling needs to take into account not just the properties of the reservoir but the cap rock as

well. Using a constant matrix property obliterates any variation in the AVO response due to

differential amounts of clay in the reservoir. Also setting input porosity same as the output

porosity removes the effect of pore pressure (mechanical compaction) which is also an

important parameter that may give rise to AVO anomalies.

However, despite these simplistic assumptions and limitations the following conclusions can

be arrived at;

The Vp depth trends in the Goliat Field generally increase with depth, except for some

anomalous zones. These zones show decreasing Vp with depth due to a combined

influence of pore fluid, pore pressure and the presence of source rocks. The chemical

compaction trend shows a smaller increase of Vp with depth (less sensitive to

increasing compaction) due to enhanced strength of the grain framework as a result of

cementation. Vp and density tend to show an inverse relationship with depth. The

source rock (Hekkingen Formation) tends to show a characteristic velocity inversion

with depth even though it has been affected by quartz cementation based on the well

log data. Sand and shale tend to show very different compaction trends. This is most

likely due to differences in the grains size.

The minimum estimate of the exhumation experienced by the Goliat Field is 700m

based on experimental laboratory compaction Kaolinite-silt (50:50) curve. Meanwhile

an upper limit based on pure kaolinite stands at 1500m based on a pure kaolinite

laboratory compaction curve. Different estimates are obtained between these two

limits depending on the laboratory compaction curve used. However XRD analysis for

the shallow formations in the mechanical compaction regime will provide the adequate

mineralogical control to constrain these results.

The present day transition from mechanical to chemical compaction occurs between

600-700m BSF, as defined by the Vp depth trend and the porosity/shear modulus cross

plot. This transition shows a clear change from mudstone to shale with a significant


92

increase in the shear modulus as a result on the onset of chemical compaction. This

temperature controlled transition occurs at similar depths in all the wells. The low

temperatures of transition estimated at 410C using a geothermal gradient of 31

0C/Km

has been attributed to a silica phase transformation, rather than a smectite to illite

transformation reaction. Future SEM petrographic analysis at depth intervals above

and below the transition zone will be important to ascertain this claim.

The Poisson’s ratio and Vp tend to be most sensitive to changing gas saturation in the

Kobbe oil-gas two phase system. Density and Vs on the other hand are only relatively

slightly sensitive. Only Vs shows an increase with increasing gas saturation due to

decreasing effective bulk density. However after 10% gas saturation, there is a

significant drop in the sensitivity of these parameters and no observed change in the

synthetic seismic. The key parameter controlling the sensitivity of Vp is the saturated

bulk modulus which is also only sensitive to the initial 10% addition gas to the full oil

system.

The insitu (real) AVO response for the Tubåen reservoir with a tight cap rock shows a

typical class IV behavior. Meanwhile the insitu (real) Kobbe AVO response shows a

class III signature. The corresponding modeled scenarios for gas show a positive ∆Vs

contribution which enhances the total increase of amplitude (becomes more negative)

with offset.

Block sizes greater than 25m average out thin reservoir intervals such as the Tubåen

reservoir. A block size of 15m has the potential to separate both cap rock and reservoir

and hence has a better resolution. However this resolution in the synthetic seismic

cannot be feasibly achieved in a real seismic survey, due to wave dispersion and

attenuation effects.

The exhumation estimates obtained can be used to correct porosity depth relationships used in

basin modeling and reservoir characterization workflows.Meanwhile AVO modeling may be

used as a complimentary tool with 3D seismic in reservoir monitoring during production

since the effects of changing saturation are less visible on the stacked seismic data compared

to the changes in the reflection coefficient as a function of offset in the prestack domain.

References

93

REFERENCES

AASE, N. E., BJORKUM, P. A. & NADEAU, P. H. 1996. The effect of grain-coating

microquartz on preservation of reservoir porosity. AAPG Bulletin-American Association of

Petroleum Geologists, 80, 1654-1673.

AKI, K. & RICHARDS, P. G. 1980. Quantitative seismology: theory and methods, San

Francisco, Freeman.

AVSETH, P., DRÆGE, A., AART-JAN, V. W., TOR, A. J. & ARILD, J. 2008. Shale rock

physics and implications for AVO analysis: A North Sea demonstration. The Leading Edge,

277, 788.

AVSETH, P., MUKERJI, T., MAVKO, G. & DVORKIN, J. 2010. Rock-physics diagnostics

of depositional texture, diagenetic alterations, and reservoir heterogeneity in high-porosity

siliciclastic sediments and rocks --- A review of selected models and suggested work flows.

Geophysics, 75, 75A31-75A47.

BATZLE, M. L., HAN, D. & CASTAGNA, J. P. 2005. Fluid effects on bright spot and AVO

analysis. SEG Technical Program Expanded Abstracts, 14, 1119 (1995).

BJØRLYKKE, k. & JAREN, H. 2010. ‘’Sandstones and Sandstone reservoirs’’. In

BJØRLYKKE, K. 2010. Petroleum Geoscience: from Sedimentary Environments to Rock

Physics, Berlin, Heidelberg, Springer-Verlag Berlin Heidelberg, 113 – 140.

CASTAGNA, J. P., BATZLE, M. L. & EASTWOOD, R. L. 1985. Relationships between

Compressional-Wave and Shear-Wave Velocities in Clastic Silicate Rocks. Geophysics, 50,

571-581.

CASTAGNA, J. P., HERBERT, W. S. & DOUGLAS, J. F. 1998. Framework for AVO

gradient and intercept interpretation. Geophysics, 63, 948-956.

DALLAND , A., WORSLEY, D. & OFSTAD, K. 1988. A lithostratigraphic scheme for the

Mesozoic and Cenoizoic succession offshore mid- and northern Norway. NPD-Bulletin, 4, 65.

DORE, A. G. 1995. Barents Sea geology ,petroleum resources and commercial potential.

Arctic, 48, 207-221.

References

94

DVORKIN, J. 2008. Yet another Vs equation. Geophysics, 73(2), E35-E39.

Eni Norway Goliat Factpage,(http://www.eninorge.no/EniNo.nsf/page

/DED71D42177627E0C12574E60040DAF9?OpenDocument&Lang=english),accessed

(02.02.11).

FALEIDE, J. I., GUDLAUGSSON, S. T. & JACQUART, G. 1984. Evolution of the western

Barents Sea. Marine and Petroleum Geology, 1, 123-128, IN1-IN4, 129-136, IN5-IN8, 137-

150.

FAWAD, M., MONDOL, N. H., JAHREN, J. & BJØRLYKKE, K. 2010. Microfabric and

rock properties of experimentally compressed silt-clay mixtures. Marine and Petroleum

Geology, 27, 1698-1712.

GABRIELSEN, R. H., FÆRSETH, R. B., JENSEN, L. N., KALHEIM, J. E. & RIIS, F. 1990.

Structural elements of the Norwegian Continental Shelf Part I : The Barents Sea Region.

NPD-Bulletin, 6, 33.

GASSMANN, F. 1951. Uber Die elastizitat poroser medien. Vier, der Natur Gesellschaft.

GELIUS, L.-J. & JOHANSEN, T. A. 2010. Petroleum geophysics, Bergen, Unigeo.

GLØRSTAD-CLARK, E., FALEIDE, J. I., LUNDSCHIEN, B. A. & NYSTUEN, J. P. 2010.

Triassic seismic sequence stratigraphy and paleogeography of the western Barents Sea area.

Marine and Petroleum Geology, 27, 1448-1475.

GOFFREY, R. 2007. Advances in reservoir characterization. Exploration and Production.

HAN, D., NUR, A. & MORGAN, D. 1986. Effects of porosity and clay content on the wave

velocities in sandstones. Geophysics, 51(11), 2093-2107.

JAPSEN, P. 1999. Overpressured Cenozoic shale mapped from velocity anomalies relative to

a baseline for marine shale, North Sea. Petroleum Geoscience, 5, 321-336.

JARVIS, K. 2006. Intergrating Well and Seismic Data for Reservoir Characterisation: Risks

and Rewards.

JOHANSEN, S. E., OSTISTY, B. K., BIRKELAND, Ø., CRISTENSEN, O., CHEREDEEV,

S. I., IGNATENKO, E. A. & MARGULIS 1993. Hydrocarbon potential in the Barents Sea

region: Play distribution and potential

http://www.eninorge.no/EniNo.nsf/page%20/DED71D42177627E0C12574E60040DAF9?OpenDocument&Lang=english

http://www.eninorge.no/EniNo.nsf/page%20/DED71D42177627E0C12574E60040DAF9?OpenDocument&Lang=english

References

95

Norwegian Petroleum Directorate (NPD) Factpages, PDO approved,

(http://factpages.npd.no/factpages/Default.aspx?culture=en) accessed (02.02.11).

Norwegian Petroleum Society (NPF) Special Publication 2, 273–320.

KRIEF, M., GARAT, J., STELLINGWERFF, J. & VENTRE, J. 1990. A petrophysical

interpretation using the velocities of P and S waves. The Log Analyst, 355-369.

LARS, H., KARSTEN, M. & ALSING, S. 2006. Improving AVO modeling using geological

knowledge, 4 examples from the Norwegian Continental Shelf. SEG Technical Program

Expanded Abstracts, 25, 259.

LOENG, H. 1991. Features of the physical oceanographic conditions of the barents sea. Polar

Research, 10, 5-18.

MARCUSSEN, O., MAAST, T. E., MONDOL, N. H., JAHREN, J. & BJØRLYKKE, K.

2010. Changes in physical properties of a reservoir sandstone as a function of burial depth -

The Etive Formation, northern North Sea. Marine and Petroleum Geology, 27, 1725-1735.

MARCUSSEN, O., THYBERG, B. I., PELTONEN, C., JAHREN, J., BJØRLYKKE, K. &

FALEIDE, J. I. 2009. Physical properties of Cenozoic mudstones from the northern North

Sea: Impact of clay mineralogy on compaction trends. AAPG Bulletin, 93, 127-150.

MOHAGHEGH, S., AREFI, R., AMERI, S., AMINIAND, K. & NUTTER, R. 1996.

Petroleum reservoir characterization with the aid of artificial neural networks. Journal of

Petroleum Science and Engineering, 16, 263-274.

MONDOL, N. H. (2011). Personal communication, University of Oslo, Norway.

MONDOL, N. H. 2010. ‘’Seismic Exploration’’. In BJØRLYKKE, K. 2010. Petroleum

Geoscience: from Sedimentary Environments to Rock Physics, Berlin, Heidelberg, Springer-

Verlag Berlin Heidelberg, 375 – 402.

MONDOL, N. H., AVSETH, P., FAWAD, M. & SMITH, T. 2010. Vs Prediction in

Unconsolidated Sands - Physical and Geological Controls on Shear Wave Velocities. 72nd

EAGE Conference & Exhibition incorporating SPE EUROPEC Barcelona, Spain.

http://factpages.npd.no/factpages/Default.aspx?culture=en

References

96

MONDOL, N. H., BJØRLYKKE, K. & JAHREN, J. 2008. Experimental compaction of

Kaolinite Aggregates: Effect of Grain Size on Mudrock Properties. EAGE Extended Abstract,

I037.

MONDOL, N. H., BJØRLYKKE, K. & JAHREN, J. 2008a. Experimental compaction of

clays: relationship between permeability and petrophysical properties in mudstones.

Petroleum Geoscience, 14, 319-337.

MONDOL, N. H., BJØRLYKKE, K., JAHREN, J. & HOEG, K. 2007. Experimental

mechanical compaction of clay mineral aggregates - Changes in physical properties of

mudstones during burial. Marine and Petroleum Geology, 24, 289-311.

MONDOL, N. H., FAWAD, M., GRANDE, L., JAHREN, J. & BJØRLYKKE, K. 2009a. Vs

Estimation in Mudrocks - Implications about Mineralogy and Grain Size. 71st EAGE

Conference and Exhibition. Amsterdam,The Netherlands.

MONDOL, N. H., FAWAD, M., GRANDE, L., JAHREN, J. & BJØRLYKKE, K. 2009b. Vs

estimation in mudrocks - Implications about mineralogy and grain size. EAGE Extended

Abstract, Z012.

MØRK, A., EMBRY, A., F. & WEITSCHAT, W. Year. Triassic transgressive-regressive

cycles in the Sverdrup Basin, Svalbard and the Barents Shelf. In: Norwegian Society

conference, 1989 Graham and Trotman,London. 113-130.

MØRK, A., KNARUD , R. & WORSLEY, D. 1982. Depositional and diagenetic

environments of the Triassic and Lower Jurassic succession of Svalbard. Canadian Society of

Geologist,Calgary Alberta, Canada, 371e398.

OHM, S. E. & KARLSEN, D. A. 2008. Geochemically driven exploration models in uplifted

areas : Examples from the Norwegian Barents Sea. AAPG Bulletin, 92,NO.9, 1191-1223.

RIIS, F., LUNDSCHIEN, B. A., HOY, T., MORK, A. & MORK, M. B. E. 2008. Evolution of

the Triassic shelf in the northern Barents Sea region. Polar Research, 27, 318-338.

ROALDSET, E. & HE, W. 1995. Silica-phase Transformation of Opal-A to Opal-CT to

Quartz-an experimental Approach.Report.: Department of Geology and Mineral Resources

Engineering, Norwegian Insitute of Technology, Trondheim.

References

97

RUTHERFORD, S. R. & WILLIAMS, R. H. 1989. AMPLITUDE-VERSUS-OFFSET

VARIATIONS IN GAS SANDS. Geophysics, 54, 680-688.

SOLHEIM, A. & ELVERHOI, A. 1993. GAS-RELATED SEA-FLOOR CRATERS IN THE

BARENTS SEA. Geo-Marine Letters, 13, 235-243.

STAINFORTH, J. G. & REINDERS, J. E. A. 1990. Primary migration of hydrocarbons by

diffusion through organic-matter networks, and its effect on oil and gas generation. Organic

Geochemistry, 16, 61-74.

STORVOLL, V., BJORLYKKE, K. & MONDOL, N. H. 2005. Velocity-depth trends in

mesozoic and cenozoic sediments from the Norwegian shelf. AAPG Bulletin, 89, 359-381.

THYBERG, B., JAHREN, J., WINJE, T., BJORLYKKE, K., FALEIDE, J. I. &

MARCUSSEN, O. 2010. Quartz cementation in Late Cretaceous mudstones, northern North

Sea: Changes in rock properties due to dissolution of smectite and precipitation of micro-

quartz crystals. Marine and Petroleum Geology, 27, 1752-1764.

VERNIK, L. & LIU, X. Z. 1997. Velocity anisotropy in shales: A petrophysical study.

Geophysics, 62, 521-532.

WALDERHAUG, O. 1994. Precipitation rates for quartz cement in sandstones determined by

fluid-inclusion microthermometry and temperature-history modeling. Journal of Sedimentary

Research Section a-Sedimentary Petrology and Processes, 64, 324-333.

WORSLEY, D. 2008. The post-Caledonian development of Svalbard and the western Barents

Sea Polar Research 27, 298 -317

WORSLEY D., AGA O.J., DALLAND A., A., E. & A., T. 1986. The geological history of

Svalbard- evolution of an Arctic archipelago

XU, S. & WHITE, R. E. 1996. A physical model for shear-wave velocity prediction.

Geophysical Prospecting, 44(4), 687-717.

YONGYI, L., JONATHAN, D. & YONG, X. 2007. Practical aspects of AVO modeling. The

Leading Edge, 26(3), 295.

Appendix

98

APPENDIX

LIST OF FIGURES

Chapter 1 Introduction

Fig. 1.1 Barents Sea exploration activity. ..................................................................................3

Fig. 1.2 Location map of Goliat Field (NPD Factpages)............................................................4

Fig. 1.3 Well locations superimposed on the outline of the Goliat Field (NPD

Factpages)....................................................................................................................7

Chapter 2 Regional Geologic Setting

Fig. 2.1 Map showing the Goliat Field and the Troms-Finnmark Fault Complex (modified

from NPD Factpages). Bathymetric map modified from Jacobsson et al.

(2008)............................................................................................................................9

Fig. 2.2 Main Structural Elements in the Barents Sea (Faleide et al., 2008, Gabrielsen et al.,

1990, Gudlaugsson et al., 1998)................................................................................. 10

Fig. 2.3 Regional stratigraphy of the Barents Shelf. The cored interval in the Nordkapp basin

are shown (modified after Bugge et al., 2002). .........................................................12

Fig. 2.4 Lithostratigraphy of the Triassic in the Western Barents Sea (modified after Glørstad-

Clark et al., 2010)........................................................................................................13

Fig. 2.5 Major source and reservoir rocks in the Barents Sea area (adapted from Dore, 1995)..

...................................................................................................................................16

Fig. 2.6 Core description of the Hekkingen Formation (adapted from Bugge et al., 2002).....17

Fig. 2.7 Core description of the Fruholmen Formation (adapted from Bugge et al., 2002).....19

Fig. 2.8 Core description of the Snadd Formation (adapted from Bugge et al., 2002).............20

Appendix

99

Fig. 2.9 Core description of the Kobbe Formation (adapted from Bugge et al., 2002)............20

Fig. 2.10 Base cretaceous unconformity depth structure map. Seismic profile shows thinner

and more faulted cap rocks (Jurassic and Cretaceous), in the Goliat area than farther

out in the basin (modified from Ohm and Karlsen, 2008).........................................22

Fig. 2.11 Gamma ray log (API) for well 7122/7-3 showing some source rock intervals (circled

in red) including cap rock horizons. Upward coarsening and fining sequences are

shown with arrows. ...................................................................................................22

Fig. 2.12 Correlation between hydrocarbon phase and cap-rock quality (adapted from Ohm

and Karlsen, 2008)....................................................................................................................23

Chapter 3 Compaction and Rock Properties

Fig. 3.1 Plots of petrophysical and acoustic properties of brine-saturated kaolinite aggregates

as a function of vertical effective stress (adapted from Mondol et al., 2008)............27

Fig. 3.2 Effect of sand grain size on mechanical compaction with increasing stress (adapted

from Bjørlykke and Jahren, 2010)...............................................................................27

Fig. 3.3 Quartz cement formation in sandstones and grain coatings (Bjørlykke and Jahren,

2010)…………………………….............................................................…….........29

Fig. 3.4 Grain coating by chlorite, well 6506/12-10, depth 5024.50m RKB, Smørbukk Field,

Haltenbanken. Adapted from (Bjørlykke and Jahren, 2010).....................................30

Fig. 3.5 Diagenesis as a function of temperature and time (adapted from Bjørlykke and

Jahren, 2010).............................................................................................................30

Fig. 3.6 Compaction trends observed in wells 7122/7-1, 7122/7-2 and 7122/7-3....................34

Fig. 3.7 Compaction trends observed in wells 7122/7-4, 7122/7-5 and 7122/7-5A. ...............35

Fig. 3.8 Well 7122/7-3 Vp-depth trend and anomalous zones.................................................37

Fig. 3.9 Shear modulus-Porosity cross plot color coded with Vshale and depth....................38

Appendix

100

Fig. 3.10 Composite shale trend compared with clay –clay and clay –silt curves...................39

Fig. 3.11 Exhumation estimates using clay –clay and clay –silt curves..................................40

Fig. 3.12 Well 7122/7-3 Vp-depth trend before and after exhumation showing transition from

mechanical to chemical compaction........................................................................41

Fig. 3.13 Complete composite well data before and after exhumation...................................41

Fig. 3.14 Variations in sand and shale compaction trends......................................................42

Fig. 3.15 Shale Vp/bulk density/porosity-depth trends...........................................................43

Fig. 3.16 Well 7122/7-3 showing possible overpressure effects. ...........................................44

Fig. 3.17 Gamma, Vp, deep resistivity and bulk density petrophysical logs for the source rock

interval (Hekkingen formation)................................................................................45

Fig. 3.18 Source rock velocity inversion..................................................................................46

Fig. 3.19 7122/7-3 anomalous zones and corresponding petrophysical logs............................48

Fig. 3.20 Tentative uplift map based on Vitrinite reflectance data (modified from Ohm and

Karlsen, 2008)...........................................................................................................51

Fig. 3.21 Subsidence curves for different regions on the Barents shelf (adapted from Ohm and

Karlsen, 2008)...........................................................................................................52

Fig. 3.22 Mineralogy of well 7117/9-1 (Roaldset and He, 1995)………….…………………54

Chapter 4 AVO Modeling

Fig. 4.1 Simple two layer model with contrasts in acoustic impedance (Z).............................61

Fig. 4.2 Convolution between the wavelet and the reflectivity series (adapted from Byørlykke

and Jahren, 2010)........................................................................................................62

Fig. 4.3 Mode conversion of P-waves (modified from Mondol, 2010)....................................63

Appendix

101

Fig. 4.4 Rutherford and Williams (1989) classification scheme based on the AVO

intercept........…………………………......................................................………....65

Fig. 4.5 AVO intercept and AVO gradient crossplot classification of gas sands (Rutherford

and Williams, 1989).....................................................................................................66

Fig. 4.6 Gamma, Deep Resistivity, Density, Vs, Vp and Computed Poisson log intervals for

the target zones (highlighted in yellow) in the Tubåen and Kobbe reservoir (well

7122/7-3)….................................................................…………………………........69

Fig. 4.7 Time and frequency domain of the Ricker linear wavelet used in this study.............70

Fig. 4.8 Resolution differences using output ‘’reflectivity’’ compared with output

‘’amplitude’’ for Kobbe reservoir in well 7122/7-3....................................................71

Fig. 4.9 Density, Vs and Vp logs generated by 15 m averaging for Tubåen (A) and Kobbe (B)

reservoirs.....................................................................................................................72

Fig. 4.10 Effect of changing gas saturations on Vp, density, Vs and Poisson’s ratio...............73

Fig. 4.11 Synthetic seismic generated using Ricker linear wavelet..........................................74

Fig. 4.12 Effect of changing gas saturations on the zero-offset reflectivity (Rp) in an oil-gas

system.........................................................................................................................75

Fig. 4.13 Angle dependent reflectivity for Tubåen Reservoir in an oil-gas system..................76

Fig. 4.14 Quantitative changes in Rp for the Tubåen reservoir after fluid replacement

modeling ...................................................................................................................77

Fig. 4.15 Kobbe reservoir offset dependent reflectivity before and after fluid substitution.....78

Fig. 4.16 Quantitative changes in Rp for the Kobbe after fluid replacement modeling...........79

Fig. 4.17 Variation in synthetic NMO corrected CMP gathers with block size 15 and 25......79

Fig. 4.18 Effect of block size on the AVO gradient.................................................................80

Fig 4.19 Gas scenario reflectivity versus offset for top Tubåen and Kobbe reservoir..............81

Appendix

102

Fig. 4.20 Cross plot between the saturated bulk modulus and density color coded with gas

saturation....................................................................................................................82

Fig. 4.21 Change in Poisson’s ratio for Tubåen reservoir after fluid substitution....................84

Fig. 4.22 Change in Poisson’s ratio for Kobbe reservoir after fluid substitution.....................87

Fig. 4.23 NMO corrected synthetic CDP gathers for block 25 and block 15 for top Tubåen

reservoir in well 7122/7-3........................................................................................................88

Fig. 4.24 Gamma ray logs for Tubåen (A) and Kobbe (B) reservoirs….................................89

LIST OF TABLES

Chapter 1 General Introduction

Table 1.1 Well data and status (modified from NPD website)................................................ 6

Chapter 2 Regional Geologic Setting

Table 2.1 Well bores and corresponding oldest Group and Formations penetrated............... 13

Table 2.2 Formations and Groups encountered in well 7122/7-3 (NPD fact pages).............. 14

Chapter 4 AVO Modeling

Table 4.1 Summary Rutherford and Williams classification scheme assuming a

‘’background’’ trend with a negative slope (Castagna et al., 1998)...…………………...…...67

Table 4.2 Kobbe formation depth and thickness variation. .................................................... 68

Table 4.3 Tubåen formation depth and thickness variation..................................................... 68

Table 4.4 Matrix properties used for fluid replacement modeling............................................71

Table 4.5 Fluid properties used for fluid replacement modeling..............................................71

Table 4.6 Variation in Vp, Vs, density and Poisson’s ratio with changing gas saturations......73

Table 4.7 Tubåen Reservoir AVO classification......................................................................76

Table 4.8 Kobbe Reservoir AVO classification. .....................................................................78

Appendix

103

Table 4.9 Density and velocity cross plot with increasing gas saturation................................82

Table 4.10 Change in Vs for Tubåen reservoir.........................................................................85

Table 4.11 Change in Vs for Kobbe reservoir..........................................................................87

compaction, rock properties and avo modeling in the … · finds is progressively decreasing. as...

Documents